Computer Science and Machine Learning Conferences [2003-2011, ENG] torrent

Обучающие видео » Компьютерные видеоуроки и обучающие интерактивные DVD » Программирование (видеоуроки)

Computer Science and Machine Learning Conferences

Год выпуска: 2003-2011
Производитель: videolectures.net
Сайт производителя: http://videolectures.net/
Тип раздаваемого материала: Видеоклипы
Язык: Английский
Продолжительность: в сумме 229 часов
Файлы примеров: не предусмотрены
Формат видео: flv

Описание: Видео выступлений с различных конференций по computer science и machine learning
[spoiler="Конференции"]
[spoiler="Machine Learning Summer School (MLSS), Tübingen 2003"]
[spoiler="Empirical Inference by Vladimir Vapnik, 2007 (rec 2003)"]
Ссылка: http://videolectures.net/mlss03_vapnik_ei
Видео: vp6f, yuv420p, 384x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Some Mathematical Tools for Machine Learning by Chris Burges, 2007 (rec 2003)"]
Описание: These are lectures on some fundamental mathematics underlying many approaches and algorithms in machine learning. They are not about particular learning algorithms; they are about the basic concepts and tools upon which such algorithms are built. Often students feel intimidated by such material: there is a vast amount of "classical mathematics", and it can be hard to find the wood for the trees. The main topics of these lectures are Lagrange multipliers, functional analysis, some notes on matrix analysis, and convex optimization. I've concentrated on things that are often not dwelt on in typical CS coursework. Lots of examples are given; if it's green, it's a puzzle for the student to think about. These lectures are far from complete: perhaps the most significant omissions are probability theory, statistics for learning, information theory, and graph theory. I hope eventually to turn all this into a series of short tutorials. Please let me know of any errors, etc. ; :
from Chris Burges homepage :
http://research.microsoft.com/~cburges
Lecture contains:
Lagrange multipliers: * Lagrange the Mathematician * Lagrange multipliers: an indirect approach can be easier * Multiple Equality Constraints * Multiple Inequality Constraints * Two points on a d-sphere * The Largest Parallelogram * Resource allocation * A convex combination of numbers is maximized by choosing the largest * The Isoperimetric problem * For fixed mean and variance, which univariate distribution has maximum entropy? * An exact solution for an SVM living on a simplex Notes on some Basic Statistics * Probabilities can be Counter-Intuitive (Simpson's paradox; the Monty Hall puzzle) * IID-ness: Measurement Error decreases as 1/sqrt{n} * Correlation versus Independence * The Ubiquitous Gaussian:
Product of Gaussians is Gaussian
Convolution of two Gaussians is a Gaussian
Projection of a Gaussian is a Gaussian
Sum of Gaussian random variables is a Gaussian random variables
Uncorrelated Gaussian variables are also independent
Maximum Likelihood Estimates for mean and covariance (prove required matrix identities)
Aside: For 1-dim Laplacian, max. likelihood gives the median * Using cumulative distributions to derive densities Principal Component Analysis and Generalizations * Ordering by Variance * Does Grouping Change Things? * PCA Decorrelates the Samples * PCA gives Reconstruction with Minimal Mean Squared Error * PCA preserves Mutual Information on Gaussian data * PCA directions lie in the span of the data * PCA: second order moments only * The Generalized Rayleigh Quotient
Non-orthogonal principal directions
OPCA
Fisher Linear Discriminant
Multiple Discriminant Analysis Elements of Functional Analysis * High Dimensional Spaces * Is Winning Transitive? * Most of the Volume is Near the Surface: Cubes * Spheres in n-dimensions * Banach Spaces, Hilbert Spaces, Compactness * Norms * Useful Inequalities (Minkowski and Holder) * Vector Norms * Matrix Norms * The Hamming Norm * L1, L2, L_infty norms - is L0 a norm? * Example: Using a Norm as a Constraint in Kernel Algorithms
Ссылка: http://videolectures.net/mlss03_burges_smtml
Видео: vp6f, yuv420p, 384x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="Machine Learning Summer School (MLSS), Berder Island 2004"]
[spoiler="Markov Chain Monte Carlo Methods by Christian Robert, 2007 (rec 2004)"]
Описание: 0. A fundamental theorem of simulation
1. Markov chain basics
2. Slice sampling
3. Gibbs sampling
4. Metropolis-Hastings algorithms
5. Variable dimension models and reversible jump MCMC
6. Perfect sampling
7. Adaptive MCMC and population Monte Carlo
Ссылка: http://videolectures.net/mlss04_robert_mcmcm
Видео: vp6f, yuv420p, 384x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Advanced Statistical Learning Theory by Olivier Bousquet, 2007 (rec 2004)"]
Описание: This set of lectures will complement the statistical learning theory course and focus on recent advances in the domain of classification. 1- PAC Bayesian bounds: a simple derivation, comparison with Rademacher averages.
2 - Local Rademacher complexity with classification loss, Talagrand's inequality. Tsybakov noise conditions.
3 - Properties of loss functions for classification (influence on approximation and estimation, relationship with noise conditions).
4 - Applications to SVM - Estimation and approximation properties, role of eigenvalues of the Gram matrix.
Ссылка: http://videolectures.net/mlss04_bousquet_aslt
Видео: vp6f, yuv420p, 352x288, 512 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[spoiler="Statistical Learning Theory by John Shawe-Taylor, 2007 (rec 2004)"]
Ссылка: http://videolectures.net/mlss04_taylor_slt
Видео: vp6f, yuv420p, 384x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="Machine Learning Summer School (MLSS), Canberra 2005"]
[spoiler="Gradient Methods for Machine Learning by Nicol Schraudolph, 2007 (rec 2005)"]
Описание: Gradient methods locally optimize an unknown differentiable function, and thus provide the engines that drive much machine learning. Here we'll take a look under the hood, beginning with brief overview of classical gradient methods for unconstrained optimization: * Steepest descent, * Newton's method * Levenberg-Marquardt * BFGS * Conjugate gradient. To cope with the flood of data we find ourselves in today, stochastic approximation of the gradient from subsamples of data becomes a necessity. Unfortunately the noise this introduces into the gradient is not tolerated well by the classical gradient methods, with the exception of steepest descent, which however is very slow to converge. We'll see how local step size adaptation can be used to accelerate the convergence of stochastic gradient descent, culminating in the recent stochastic meta-descent (SMD) algorithm. SMD requires certain Hessian-vector products which can be computed efficiently via algorithmic (or automatic) differentiation (AD), a set of techniques that help automate the correct implementation of gradient methods in general. We'll discuss the basic concepts of AD, and learn simple ways to implement the forward mode of AD, and with it the fast Hessian-vector product.
Ссылка: http://videolectures.net/mlss05au_schraudolph_gmml
Видео: vp6f, yuv420p, 384x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="Machine Learning Summer School (MLSS), Chicago 2005"]
[spoiler="Trees for Regression and Classification by Robert Nowak, 2007 (rec 2005)"]
Описание: Tree models are widely used for regression and classification problems, with interpretability and ease of implementation being among their chief attributes. Despite the widespread use tree models, a comprehensive theoretical analysis of their performance has only begun to emerge in recent years.
This lecture provides an overview of tree modeling theory and methods, with an emphasis on risk bounds, oracle inequalities, approximation theory, and rates of convergence, in a variety of contexts. Special attention is devoted to decision trees and wavelet-based regression methods, two of the most well-known examples of tree models. The choice of loss function (squared error, absolute error, 0/1 error) plays a pivotal role in both theory and methods.
In particular, optimal tree selection rules vary dramatically depending on the loss function employed. Despite these differences, suitable tree-based models coupled with appropriate selection rules can provide fast algorithms and near-minimax optimal performance in a very broad range of regression and classification problems. Examples from image reconstruction and pattern classification will demonstrate the effectiveness of trees in practice.
Ссылка: http://videolectures.net/mlss05us_nowak_trc
Видео: vp6f, yuv420p, 352x288, 512 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[spoiler="Boosting by Robert Schapire, 2007 (rec 2005)"]
Описание: Boosting is a general method for producing a very accurate classification rule by combining rough and moderately inaccurate "rules of thumb." While rooted in a theoretical framework of machine learning, boosting has been found to perform quite well empirically. This tutorial will introduce the boosting algorithm AdaBoost?, and explain the underlying theory of boosting, including explanations that have been given as to why boosting often does not suffer from overfitting, as well as some of the myriad other theoretical points of view that have been taken on this algorithm. Some recent applications and extensions of boosting will also be described.
Ссылка: http://videolectures.net/mlss05us_schapire_b
Видео: vp6f, yuv420p, 384x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Energy-based models & Learning for Invariant Image Recognition by Yann LeCun, 2007 (rec 2005)"]
Ссылка: http://videolectures.net/mlss05us_lecun_ebmli
Видео: vp6f, yuv420p, 352x288, 512 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[/spoiler]
[spoiler="Machine Learning Summer School (MLSS), Taipei 2006"]
[spoiler="Introduction to Boosting by Gunnar Rätsch, 2007 (rec 2006)"]
Описание: This course provides an introduction to theoretical and practical aspects of Boosting and Ensemble Learning. I will begin with a short description of the learning theoretical foundations of weak learners and their linear combination. Then we point out the useful connection between Boosting and the Theory of Optimization, which facilitates the understanding of Boosting and later on enables us to move on to new Boosting algorithms, applicable to a broader spectrum of problems. In the course we will discuss "tricks of the trade", algorithmic issues, experimental results and a few applications.
Ссылка: http://videolectures.net/mlss06tw_ratsch_ib
Видео: vp6f, yuv420p, 320x256, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[spoiler="Support Vector Machines by Chih-Jen Lin, 2007 (rec 2006)"]
Описание: Support vector machines (SVM) and kernel methods are important machine learning techniques. In this short course, we will introduce their basic concepts. We then focus on the training and optimization procedures of SVM. Examples demonstrating the practical use of SVM will also be discussed. Basically we focus on classification. If time is allowed, we will also touch SVM regression.
Ссылка: http://videolectures.net/mlss06tw_lin_svm
Видео: vp6f, yuv420p, 320x240, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[spoiler="Machine Learning, Probability and Graphical Models by Sam Roweis, 2007 (rec 2006)"]
Ссылка: http://videolectures.net/mlss06tw_roweis_mlpgm
Видео: vp6f, yuv420p, 320x240, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 98 kb/s
[/spoiler]
[/spoiler]
[spoiler="Gaussian Processes in Practice Workshop, Bletchley Park 2006"]
[spoiler="Gaussian Process Basics by David MacKay, 2007 (rec 2006)"]
Описание: How on earth can a plain old Gaussian distribution be useful for sophisticated regression and machine learning tasks?
Ссылка: http://videolectures.net/gpip06_mackay_gpb
Видео: vp6f, yuv420p, 320x240, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[/spoiler]
[spoiler="NIPS Workshop on Learning to Compare Examples, Whistler 2006"]
[spoiler="Neighbourhood Components Analysis and Metric Learning by Sam Roweis, 2007 (rec 2006)"]
Описание: Say you want to do K-Nearest Neighbour classification. Besides
selecting K, you also have to chose a distance function, in order to
define ”nearest”. I’ll talk about a method for learning – from the
data itself – a distance measure to be used in KNN classification. The
learning algorithm, Neighbourhood Components Analysis (NCA) directly
maximizes a stochastic variant of the leave-one-out KNN score
on the training set. Of course, the resulting classification model is
non-parametric, making no assumptions about the shape of the class
distributions or the boundaries between them. I will also discuss an
variant of the method which is a generalization of Fisher’s discriminant
and defines a convex optimization problem by trying to collapse
all examples in the same class to a single point and trying to push
examples in other classes infinitely far away. By approximating the
metric with a low rank matrix, these learning algorithms, can also be
used to obtain a low-dimensional linear embedding of the original input
features allowing that can be used for data visualization and very
fast classification in high dimensions.
Ссылка: http://videolectures.net/lce06_roweis_ncaml
Видео: vp6f, yuv420p, 320x240, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[spoiler="Learning Similarity Metrics with Invariance Properties by Yann LeCun, 2007 (rec 2006)"]
Ссылка: http://videolectures.net/lce06_lecun_lsmip
Видео: vp6f, yuv420p, 320x240, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[/spoiler]
[spoiler="Machine Learning Summer School (MLSS), Canberra 2006"]
[spoiler="Rapid Stochastic Gradient Descent: Accelerating Machine Learning by Nicol Schraudolph, 2007 (rec 2006)"]
Описание: The incorporation of online learning capabilities into real-time computing systems has been hampered by a lack of efficient, scalable optimization algorithms for this purpose: second-order methods are too expensive for large, nonlinear models, conjugate gradient does not tolerate the noise inherent in online learning, and simple gradient descent, evolutionary algorithms, etc., are unacceptably slow to converge. I am addressing this problem by developing new ways to accelerate stochastic gradient descent, using second-order gradient information obtained through the efficient computation of curvature matrix-vector products. In the stochastic meta-descent (SMD) algorithm, this cheap curvature information is built up iteratively into a stochastic approximation of Levenberg-Marquardt second-order gradient steps, which are then used to adapt individual gradient step sizes. SMD handles noisy, correlated, non-stationary signals well, and approaches the rapid convergence of second-order methods at only linear cost per iteration, thus scaling up to extremely large nonlinear systems. To date it has enabled new adaptive techniques in computational fluid dynamics and computer vision. Our most recent development is a version of SMD operating in reproducing kernel Hilbert space.
Ссылка: http://videolectures.net/mlss06au_schraudolph_aml
Видео: vp6f, yuv420p, 320x240, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[/spoiler]
[spoiler="NIPS Workshop on Dynamical Systems, Stochastic Processes and Bayesian Inference, Whistler 2006"]
[spoiler="A Tutorial Introduction to Stochastic Differential Equations: Continuous-time Gaussian Markov Processes by Chris Williams, 2007 (rec 2006)"]
Ссылка: http://videolectures.net/dsb06_williams_ctgmp
Видео: vp6f, yuv420p, 320x240, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[/spoiler]
[spoiler="24th Annual International Conference on Machine Learning (ICML), Corvallis 2007"]
[spoiler="Best Paper - Information-Theoretic Metric Learning by Brian Kulis, 2007"]
Описание: In this paper, we present an information-theoretic approach to learning a Mahalanobis distance function. We formulate the problem as that of minimizing the differential relative entropy between two multivariate Gaussians under constraints on the distance function. We express this problem as a particular Bregman optimization problem: that of minimizing the LogDet divergence subject to linear constraints. Our resulting algorithm has several advantages over existing methods. First, our method can handle a wide variety of constraints and can optionally incorporate a prior on the distance function. Second, it is fast and scalable. Unlike most existing methods, no eigenvalue computations or semi-definite programming are required. We also present an online version and derive regret bounds for the resulting algorithm. Finally, we evaluate our method on a recent error reporting system for software called Clarify, in the context of metric learning for nearest neighbor classification, as well as on standard data sets.
Ссылка: http://videolectures.net/icml07_kulis_itml
Видео: vp6f, yuv420p, 320x256, 409 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[/spoiler]
[spoiler="Machine Learning Summer School (MLSS), Tübingen 2007"]
[spoiler="Introduction to kernel methods by Alexander Smola, Bernhard Scholkopf, 2007"]
Описание: This lecture given by Mr. Smola is combined with Mr. Bernhard Schoelkopf and will encopass Part 1, Part 5, Part 6 of the complete lecture.
Part 2, 3 and 4 of this lecture can be found here at Bernhard Schoelkopf's "Introduction to kernel methods".

This lecture given by Mr. Bernhard Scholkop is combined with Mr. Smola and will encompass Part 2, Part 3, Part 4 of the complete lecture. Part 1 , 5, 6 of this lecture can be found here at Alex Smola's "Introduction to kernel methods".
Ссылка: http://videolectures.net/mlss07_smola_intkmet
http://videolectures.net/mlss07_scholkopf_intkmet
Видео: vp6f, yuv420p, 320x256, 409 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[/spoiler]
[spoiler="The Analysis of Patterns, Bertinoro 2007"]
[spoiler="Support Vector Machines and Kernel Methods by Colin Campbell, 2007"]
Ссылка: http://videolectures.net/aop07_campbell_svm
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="NIPS Workshop on Efficient Machine Learning, Whistler 2007"]
[spoiler="New Quasi-Newton Methods for Efficient Large-Scale Machine Learning by S.V.N. Vishwanathan, 2007"]
Описание: The BFGS quasi-Newton method and its limited-memory variant LBFGS revolutionized nonlinear optimization, and dominate it to this day. Their application to large-scale machine learning, however, has been hindered by the fact that they assume a smooth, strictly convex, and deterministic objective function in a finite-dimensional vector space. Here we relax these assumptions one by one, and present (L)BFGS variants newly developed in our group that perform well on non-convex smooth, quasi-convex non-smooth, and non-deterministic objectives. Paradigmatic applications include parameter estimation in MLPs (non-convex smooth) and SVMs (convex non-smooth), and stochastic approximation of gradients (non-deterministic) for efficient online learning on large data sets.
We are also able to lift LBFGS to an RKHS for online SVM training. In all these cases our BFGS variants outperform previous methods on a wide variety of models and data sets, from toy problems to large-scale data-mining tasks.
Ссылка: http://videolectures.net/eml07_vishwanathan_nqm
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Speeding Up Stochastic Gradient Descent by Yoshua Bengio, 2007"]
Описание: n order to tackle large-scale learning problems whose solution necessarily involves a large model with many tunable parameters, difficult non-convex optimization has to be performed efficiently. Computational complexity arguments strongly suggest that deep architectures will be necessary to represent the kind of complex functions that AI involves. Unfortunately, this involves difficult optimization problems and efficient approximate iterative optimization becomes key to obtain good generalization, and not so much the regularization techniques that have been so well studied in the last two decades. Furthermore, because of the size of the data sets involved in such tasks, it is imperative that computation scale no more than linearly with respect to the number of training examples. In many cases, the algorithm to beat is stochastic gradient descent, and the comparisons have to be made by looking at the curve of test error versus computation time. Following recent interest in online versions of second-order optimization methods, we present computational tricks that yield a linear time variant of natural gradient optimization. Another issue, that is particularly difficult to address in the optimization of multi-layer neural networks, is how to parallelize efficiently. SMP machines becoming cheaper and easier to use, we compare and discuss different strategies for exploiting parallelization of training for multi-layer neural networks, showing that naive approaches fail but those taking into account the communication bottleneck yield impressive speed-ups.
Ссылка: http://videolectures.net/eml07_bengio_ssg
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Who is Afraid of Non-Convex Loss Functions? by Yann LeCun, 2007"]
Описание: The NIPS community has suffered of an acute convexivitis epidemic:
- ML applications seem to have trouble moving beyond logistic regression, SVMs, and exponential-family graphical models;
- For a new ML model, convexity is viewed as a virtue;
- Convexity is sometimes a virtue;
- But it is often a limitation.
ML theory has essentially never moved beyond convex models - the same way control theory has not really moved beyond linear systems.
Ссылка: http://videolectures.net/eml07_lecun_wia
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Interview with Yann LeCun by Yann LeCun, 2008 (rec 2007)"]
Описание: His lab has projects in computer vision, object detection, object recognition, mobile robotics, bio-informatics, biological image analysis, medical signal processing, signal processing, and financial prediction,...The Videolectures.Net team talked to him at NIPS 2007, we asked him stuff like: *What is your current topic of research? *How can you comment on your humoristic approach in giving lectures? *Humor and content? *What happend in your research between ML Summer school 2006 in Chicago and today? *How can you explain your work to a 4 year old child? *Machine Learning dream come true.... *What is your philosophy?
Ссылка: http://videolectures.net/eml07_lecun_int
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="PASCAL Bootcamp in Machine Learning, Vilanova 2007"]
[spoiler="Basics of probability and statistics by Mikaela Keller, 2007"]
Ссылка: http://videolectures.net/bootcamp07_keller_bss
Видео: vp6f, yuv420p, 320x256, 409 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[/spoiler]
[spoiler="EPSRC Winter School in Mathematics for Data Modelling, Sheffield 2008"]
[spoiler="Introduction to Support Vector Machines by Colin Campbell, 2008"]
Описание: This first presentation introduces support vector machines.
Ссылка: http://videolectures.net/epsrcws08_campbell_isvm
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="NIPS Workshop on Optimization for Machine Learning, Whistler 2008"]
[spoiler="Large-scale Machine Learning and Stochastic Algorithms by Léon Bottou, 2008"]
Описание: The presentation stresses important differences between machine learning and conventional optimisation approaches and proposes some solutions. The first part discusses the the interaction of two kind of asympotic properties: those of the statistics and those of optimization algorithm. Unlikely optimization algorithm such as stochastic gradient show amazing performance for large-scale machine learning problems. The second part shows how the deeper causes of this performance suggests the theoretical possibility learn large-scale problems with a single pass over the data. Practical algorithms will be discussed: various second order stochastic gradients, averaging methods, dual methods with data reprocessing...
Ссылка: http://videolectures.net/opt08_bottou_lsml
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="Machine Learning Summer School (MLSS), Kioloa 2008"]
[spoiler="Introduction to Statistical Machine Learning by Marcus Hutter, 2008"]
Описание: The first part of his tutorial provides a brief overview of the fundamental methods
and applications of statistical machine learning.
The other speakers will detail or built upon this introduction.
Statistical machine learning is concerned with
the development of algorithms and techniques that learn from
observed data by constructing stochastic models that can be used for
making predictions and decisions.
Topics covered include Bayesian inference and maximum likelihood
modeling; regression, classification, density estimation,
clustering, principal component analysis; parametric,
semi-parametric, and non-parametric models; basis functions, neural
networks, kernel methods, and graphical models; deterministic and
stochastic optimization; overfitting, regularization, and
validation.
Ссылка: http://videolectures.net/mlss08au_hutter_isml
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Learning in Computer Vision by Simon Lucey, 2008"]
Описание: This tutorial he will cover some of the core fundamentals in vision and demonstrate how they can be interpreted in terms of machine learning fundamentals. Unbeknownst to most researchers in the field of machine learning, the fundamentals of object registration and tracking such as optical flow, interest descriptors (e.g., SIFT), segmentation and correlation filters are inherently related to the learning topics of regression, regularization, graphical models, generative models and discriminative models. As a result many aspects of vision can be interpreted as applied forms of learning. From this discussion on fundamentals we shall also explore advanced topics in object registration and tracking such as non-rigid object alignment/ tracking and non-rigid structure from motion and how the application of machine learning is continuing to improve these technologies.
Ссылка: http://videolectures.net/mlss08au_lucey_linv
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Foundations of Machine Learning by Marcus Hutter, 2008"]
Описание: Machine learning is usually taught as a bunch of methods that can
solve a bunch of problems (see above).
The second part of the tutorial takes a step back and asks about the
foundations of machine learning, in particular the (philosophical)
problem of inductive inference, (Bayesian) statistics, and
artificial intelligence.
It concentrates on principled, unified, and exact methods.
Ссылка: http://videolectures.net/mlss08au_hutter_fund
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Kernel methods and Support Vector Machines by Alexander J. Smola, 2008"]
Описание: The tutorial will introduce the main ideas of statistical learning
theory, support vector machines, and kernel feature spaces.
This includes a derivation of the support vector optimization
problem for classification and regression, the v-trick,
various kernels and an overview over applications of kernel methods.
Ссылка: http://videolectures.net/mlss08au_smola_ksvm
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Introduction to Reinforcement Learning by Csaba Szepesvari, 2008"]
Описание: The tutorial will introduce
Reinforcement Learning, that is, learning what actions to take,
and when to take them, so as to optimize long-term performance. This may
involve sacrificing immediate reward to obtain greater reward in the
long-term or just to obtain more information about the environment. The
first part of the tutorial will cover the basics, such as Markov
decision processes, dynamic programming, temporal-difference learning,
Monte Carlo methods, eligibility traces, the role of function
approximation. In the second part we cover some recent developments,
namely policy gradient and second order methods, such as LSPI and the
modified Bellman residual minimization algorithm.
Ссылка: http://videolectures.net/mlss08au_szepesvari_rele
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Monte Carlo Simulation for Statistical Inference, Model Selection and Decision Making by Nando de Freitas, 2008"]
Описание: The first part
of his course will consist of two presentations. In the first presentation, he will introduce fundamentals of Monte Carlo simulation for statistical inference, with emphasis on algorithms such as importance sampling, particle filtering and smoothing for dynamic models, Markov chain Monte Carlo, Gibbs and Metropolis-Hastings, blocking and mixtures of MCMC kernels, Monte Carlo EM, sequential Monte Carlo for static models, auxiliary variable methods (Swedsen-Wang, hybrid Monte Carlo and slice sampling), and adaptive MCMC. The algorithms will be illustrated with several examples: image tracking, robotics, image annotation, probabilistic graphical models, and music analysis.
The second presentation
will target model selection and decision making problems. He will describe the reversible-jump MCMC algorithm and illustrate it with application to simple mixture models and nonlinear regression with an unknown number of basis functions. He will show how to apply this algorithm to general Markov decision processes (MDPs). The course will also cover other Monte Carlo simulation methods for partially observed Markov decision processes (POMDPs) using policy gradients, common random number generation, and active exploration with Gaussian processes. An outline to some applications of these methods to robotics and the design of computer game architectures will be given. The presentation will end with the problem of Monte Carlo simulation for Bayesian nonlinear experimental design, with application to financial modeling, robot exploration, drug treatments, dynamic sensor networks, optimal measurement and active vision.
Ссылка: http://videolectures.net/mlss08au_freitas_asm
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="CERN Summer School 2009"]
[spoiler="Introduction to Statistics by Glen Cowan, 2010 (rec 2009)"]
Ссылка: http://videolectures.net/cernstudentsummerschool09_cowan_is
Видео: flv, yuv420p, 352x288, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 128 kb/s
[/spoiler]
[/spoiler]
[spoiler="Summer Schools in Logic and Learning, Canberra 2009"]
[spoiler="Reinforcement learning by Scott Sanner, 2009"]
Описание: This course covers the theory and application of reinforcement learning: the task of learning to make optimal sequential decisions when given a delayed reward signal. Topics will include planning in known and unknown environments and will place equal emphasis on theoretical results and practical implementation issues in the context of various applications.
Ссылка: http://videolectures.net/ssll09_sanner_rele
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Learning Theory by Mark Reid, 2009"]
Описание: This course highlights some relationships between surrogate losses, scoring rules, f-divergences, Bregman divergences, statistical information and ROC curves and their implications for applications such as divergence estimation.
Ссылка: http://videolectures.net/ssll09_reid_leth
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Computer vision by Richard Hartley, 2009"]
Описание: A pseudo-boolean function is a function from the space B^n of boolean (0-1) vector to the real numbers. They occur naturally in problems in computer vision related to segmentation where every pixel in an image should be labelled 0 or 1 to minimize a certain cost function. Although the minimization of such functions in in general NP hard, many techniques have been develloped to minimize certain classes of such functions. This is the topic of pseudo-boolean optimization, which will be the subject of this talk. Useful methods include graph-cuts algorithms, message passing and linear programming relaxation. The extension to functions with a finite label set will also be considered.
Ссылка: http://videolectures.net/ssll09_hartley_covi
Видео: vp6f, yuv420p, 352x288, 716 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Group Theory in Machine Learning by Marconi Barbosa, 2009"]
Описание: This course covers diverse aspects of the role played by symmetry in pattern analysis and machine learning. It is designed to provide background knowledge using examples and to touch current research topics without over emphasizing formalizations and technical descriptions.
Ссылка: http://videolectures.net/ssll09_barbosa_gtiml
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Graphical models by Tibério Caetano, 2009"]
Описание: This course covers the basics of Probabilistic Graphical Models, including the basic theory of Bayesian Networks and Markov Random Fields, as well as inference and learning algorithms and applications.
Ссылка: http://videolectures.net/ssll09_caetano_grmo
Видео: vp6f, yuv420p, 352x288, 716 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="Machine Learning Summer School (MLSS), Cambridge 2009"]
[spoiler="Approximate Inference by Tom Minka, 2009"]
Ссылка: http://videolectures.net/mlss09uk_minka_ai
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Information Theory by David MacKay, 2009"]
Ссылка: http://videolectures.net/mlss09uk_mackay_it
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Learning Theory by John Shawe-Taylor, 2009"]
Ссылка: http://videolectures.net/mlss09uk_shawe_taylor_lt
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Introduction To Bayesian Inference by Christopher Bishop, 2009"]
Ссылка: http://videolectures.net/mlss09uk_bishop_ibi
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Gaussian Processes by Carl Edward Rasmussen, 2009"]
Ссылка: http://videolectures.net/mlss09uk_rasmussen_gp
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Deep Belief Networks by Geoffrey E. Hinton, 2009"]
Ссылка: http://videolectures.net/mlss09uk_hinton_dbn
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Computer Vision by Andrew Blake, 2009"]
Ссылка: http://videolectures.net/mlss09uk_blake_cv
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Markov Chain Monte Carlo by Iain Murray, 2009"]
Ссылка: http://videolectures.net/mlss09uk_murray_mcmc
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Reinforcement Learning by Michael Littman, 2009"]
Ссылка: http://videolectures.net/mlss09uk_littman_rl
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Bayesian or Frequentist, Which Are You? by Michael I. Jordan, 2009"]
Ссылка: http://videolectures.net/mlss09uk_jordan_bfway
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="26th International Conference on Machine Learning (ICML), Montreal 2009"]
[spoiler="Tutorial on Learning Deep Architectures by Yoshua Bengio, Yann LeCun, 2009"]
Описание: This short tutorial on deep learning will review a variety of methods for learning multi-level, hierarchical representations, emphasizing their common traits. While deep architectures have theoretical advantages in terms of expressive power and efficiency of representation, they also provide a possible model for information processing in the mammalian cortex, which seems to rely on representations with multiple levels of abstractions. A number of deep learning methods have been proposed since 2005, that have yielded surprisingly good performance in several areas, particularly in vision (object recognition), and natural language processing. They all learn multiple levels of representation using some form of unsupervised learning. Hypotheses to explain why these algorithms work well will be discussed in the light of new experimental results. Many of these algorithms can be cast in the framework of the energy-based view of unsupervised learning, which generalizes graphical models used as building blocks for deep architectures, such as the Restricted Boltzmann Machines (RBM) and variations of regularized auto-encoders. Old and new algorithms will be presented for training, sampling, and estimating the partition function of RBMs and Deep Belief Networks. Applications of deep architectures to computer vision and natural language processing will be described. A number of open problems and future research avenues will be discussed, with active participation from the audience.
Ссылка: http://videolectures.net/icml09_bengio_lecun_tldar
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="A Factor Model for Learning Higher Order Features in Natural Images by Yan Karklin, 2009"]
Описание: The visual system is a hierarchy of processing stages. Each stage in this pathway, in addition to encoding increasingly complex features of the input, performs complex non-linear computations. What is the functional role of these non-linear behaviors and how do we incorporate them into generative models of natural images?
A number of non-linear properties of visual neurons can be predicted from the statistical dependencies observed in natural images. For example, the magnitudes of linear filter outputs are correlated; normalizing filter responses removes this correlation (making the responses more independent and marginally Gaussian) and reproduces neural gain control. In addition, the pattern in these correlations is itself highly informative, and can be used to infer the context of patches sampled from a large scene. Here I will focus on these statistical patterns and describe a generative model that captures them using a set of factors in the space of log-covariance of a multivariate Gaussian distribution. Trained on natural images, the model learns a compact code for correlations observed in pixel (or linear feature) distributions that represents more abstract properties of the image. I will also connect this work to recent generative models that incorporate multiplicative interactions between observed and latent variables.
Ссылка: http://videolectures.net/icml09_karklin_fmlh
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Unsupervised Discovery of Structure, Succinct Representations and Sparsity by Andrew Y. Ng, 2009"]
Описание: We describe a class of unsupervised learning methods that learn sparse representations of the training data, and thereby identify useful features. Further, we show that deep learning (multilayer) versions of these ideas, ones based on sparse DBNs, learn rich feature hierarchies, including part-whole decompositions of objects. Central to this is the idea of "probabilistic max pooling", which allows us to implement convolutional DBNs at a large scale, while maintaining probabilistically sound semantics. In the case of images, at the lowest level this method learns to detect edges; at the next level, it puts together edges to form "object parts"; and finally, at the highest level puts together object parts to form whole "object models". The features this method learns are useful for a wide range of tasks, including object recognition, text classification, and audio classification. We also present the result of comparing a two-layer version of the model (trained on natural images) to visual cortical areas V1 and V2 in the brain (the first and second stages of visual processing in the cortex). Finally, we'll conclude with a discussion on some open problems and directions for future research.
Ссылка: http://videolectures.net/icml09_ng_udssrs
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Matrix Computations in Machine Learning by Inderjit S. Dhillon, 2009"]
Описание: Matrix Computations are ubiquitous in all areas of science and engineering. In this talk, I will first survey some traditional problems in matrix computations and discuss issues that arise in solving them, such as, accuracy, algorithms and software. Then, I will discuss various matrix computation problems that arise in machine learning, especially specialized computations, such as non-negative matrix factorization, multilevel graph clustering and kernel learning. I will conclude with a pointer to resources and a discussion.
Ссылка: http://videolectures.net/icml09_dhillon_itmcml
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="Machine Learning Summer School (MLSS), Chicago 2009"]
[spoiler="Learning Dictionaries for Image Analysis and Sensing by Guillermo Sapiro, 2009"]
Описание: Sparse representations have recently drawn much attention from the signal processing and learning communities. The basic underlying model consist of considering that natural images, or signals in general, admit a sparse decomposition in some redundant dictionary. This means that we can find a linear combination of a few atoms from the dictionary that lead to an efficient representation of the original signal. Recent results have shown that learning overcomplete non-parametric dictionaries for image representations, instead of using off-the-shelf ones, significantly improves numerous image and video processing tasks.
In this talk, I will first present our results on learning multiscale overcomplete dictionaries for color image and video restoration. I will present the framework and provide numerous examples showing state-of-the-art results. I will then briefly show how to extend this to image classification, deriving energies and optimization procedures that lead to learning non-parametric dictionaries for sparse representations optimized for classification. I will conclude by showing results on the extension of this to sensing and the learning of incoherent dictionaries. The work I present in this talk is the result of great collaborations with J. Mairal (ENS, Paris), F. Rodriguez (UofM/Spain), J. Martin-Duarte (UofM/Kodak), I. Ramirez (UofM), F. Lecumberry (UofM), F. Bach (ENS, Paris), M. Elad (Technion, Israel), J. Ponce (ENS, Paris), and A. Zisserman (ENS/Oxford).
Ссылка: http://videolectures.net/mlss09us_sapiro_ldias
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Seeking Interpretable Models for High Dimensional Data by Bin Yu, 2009"]
Описание: Extracting useful information from high-dimensional data is the focus of today's statistical research and practice. After broad success of statistical machine learning on prediction through regularization, interpretability is gaining attention and sparsity has been used as its proxy. With the virtues of both regularization and sparsity, Lasso (L1 penalized L2 minimization) has been very popular recently. In this talk, I would like to discuss the theory and pratcice of sparse modeling. First, I will give an overview of recent research on sparsity and explain what useful insights have been learned from theoretical analyses of Lasso. Second, I will present collaborative research with the Gallant Lab at Berkeley on building sparse models (linear, nonlinear, and graphical) that describe fMRI responses in primary visual cortex area V1 to natural images.
Ссылка: http://videolectures.net/mlss09us_yu_simhdd
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Semi-Supervised Learning by Jerry (Xiaojin) Zhu, 2009"]
Описание: This tutorial covers classification approaches that utilize both labeled and unlabeled data. We will review self-training, Gaussian mixture models, co-training, multiview learning, graph-transduction and manifold regularization, transductive SVMs, and a PAC bound for semi-supervised learning. We then discuss some new development, including online semi-supervised learning, multi-manifold learning, and human semi-supervised learning.
Ссылка: http://videolectures.net/mlss09us_zhu_ssl
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Sparse Representations from Inverse Problems to Pattern Recognition by Stéphane Mallat, 2009"]
Описание: Sparse representations are at the core of many low-level signal processing procedures and are used by most pattern recognition algorithms to reduce the dimension of the search space. Structuring sparse representations fro pattern recognition applications requires taking into account invariants relatively to physical deformations such as rotation scaling or illumination. Sparsity, invariants and stability are conflicting requirements which is a source of open problems. Structured sparse representations with locally linear vector spaces are introduced for super-resolution inverse problems and pattern recognition.
Ссылка: http://videolectures.net/mlss09us_mallat_srippr
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Kernel Methods and Support Vector Machines by John Shawe-Taylor, 2009"]
Описание: Kernel methods have become a standard tool for pattern analysis during the last fifteen years since the introduction of support vector machines. We will introduce the key ideas and indicate how this approach to pattern analysis enables a relatively easy plug and play application of different tools. The problem of choosing and designing a kernel for specific types of data will also be considered and an overview of different kernels will be given.
Ссылка: http://videolectures.net/mlss09us_shawe-taylor_kmsvm
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="How to Visualize the Unseeable by Xiaochuan Pan, 2009"]
Ссылка: http://videolectures.net/mlss09us_pan_hvu
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Fitting a Graph to Vector Data by Daniel A. Spielman, 2009"]
Описание: We ask "What is the right graph to fit to a set of vectors?" We propose one solution that provides good answers to standard Machine Learning problems, that has interesting combinatorial properties, and that we can compute efficiently. Joint work with Jonathan Kelner and Samuel Daitch.
Ссылка: http://videolectures.net/mlss09us_spielman_fgvd
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Matrix Completion via Convex Optimization: Theory and Algorithms by Emmanuel Candes, 2009"]
Описание: This talk considers a problem of considerable practical interest: the recovery of a data matrix from a sampling of its entries. In partially filled out surveys, for instance, we would like to infer the many missing entries. In the area of recommender systems, users submit ratings on a subset of entries in a database, and the vendor provides recommendations based on the user's preferences. Because users only rate a few items, we would like to infer their preference for unrated items (this is the famous Netflix problem). Formally, suppose that we observe m entries selected uniformly at random from a matrix. Can we complete the matrix and recover the entries that we have not seen? We show that perhaps surprisingly, one can recover low-rank matrices exactly from what appear to be highly incomplete sets of sampled entries; that is, from a minimally sampled set of entries. Further, perfect recovery is possible by solving a simple convex optimization program, namely, a convenient semidefinite program. A surprise is that our methods are optimal and succeed as soon as recovery is possible by any method whatsoever, no matter how intractable; this result hinges on powerful techniques in probability theory. Time permitting, we will also present a very efficient algorithm based on iterative singular value thresholding, which can complete matrices with about a billion entries in a matter of minutes on a personal computer.
Ссылка: http://videolectures.net/mlss09us_candes_mccota
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Optimization Algorithms in Support Vector Machines by Stephen J. Wright, 2009"]
Описание: This talk presents techniques for nonstationarity detection in the context of speech and audio waveforms, with broad application to any class of time series that exhibits locally stationary behavior. Many such waveforms, in particular information-carrying natural sound signals, exhibit a degree of controlled nonstationarity, and are often well modeled as slowly time-varying systems. The talk first describes the basic concepts of such systems and their analysis via local Fourier methods. Parametric approaches appropriate for speech are then introduced by way of time-varying autoregressive models, along with nonparametric approaches based on variation of time-localized estimates of the power spectral density of an observed random process, along with an efficient offline bootstrap procedure based on the Wold representation. Several real-world examples are given.
Ссылка: http://videolectures.net/mlss09us_wright_oasvm
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Learning Feature Hierarchies by Yann LeCun, 2009"]
Ссылка: http://videolectures.net/mlss09us_lecun_lfh
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Theory and Applications of Boosting by Robert Schapire, 2009"]
Описание: Boosting is a general method for producing a very accurate classification rule by combining rough and moderately inaccurate "rules of thumb". While rooted in a theoretical framework of machine learning, boosting has been found to perform quite well empirically. This tutorial will introduce the boosting algorithm AdaBoost, and explain the underlying theory of boosting, including explanations that have been given as to why boosting often does not suffer from overfitting, as well as some of the myriad other theoretical points of view that have been taken on this algorithm. Some practical applications and extensions of boosting will also be described.
Ссылка: http://videolectures.net/mlss09us_schapire_tab
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Unsupervised Learning for Stereo Vision by David McAllester, 2009"]
Описание: We consider the problem of learning to estimate depth from stereo image pairs. This can be formulated as unsupervised learning - the training pairs are not labeled with depth. We have formulated an algorithm which maximizes conditional likelihood the left image given right image in a model that involves latent information (depth). This unsupervised learning algorithm implicitly trains shape from texture and shape from shading monocular depth cues. The talk will present pragmatic results in the stereo vision problem as well as a general formulation of models and methods for maximizing conditional likelihood in a latent variable model where we wish to interpret the latent information as "labels".
Ссылка: http://videolectures.net/mlss09us_mcallester_ulsv
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Bounding Excess Risk in Machine Learning by Vladimir Koltchinskii, 2009"]
Описание: We will discuss a general approach to the problem of bounding the excess risk of learning algorithms based on empirical risk minimization (possibly penalized). This approach has been developed in the recent years by several authors (among others: Massart; Bartlett, Bousquet and Mendelson; Koltchinskii). It is based on powerful concentration inequalities due to Talagrand as well as on a variety of tools of empirical processes theory (comparison inequalities, entropy and generic chaining bounds on Gaussian, empirical and Rademacher processes, etc.). It provides a way to obtain sharp excess risk bounds in a number of problems such as regression, density estimation and classification and for many different classes of learning methods (kernel machines, ensemble methods, sparse recovery). It also provides a general way to construct sharp data dependent bounds on excess risk that can be used in model selection and adaptation problems.
Ссылка: http://videolectures.net/mlss09us_koltchinskii_berml
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="Machine Learning Summer School (MLSS), Canberra 2010"]
[spoiler="Online Learning by Peter L. Bartlett, 2011 (rec 2010)"]
Ссылка: http://videolectures.net/mlss2010au_bartlett_onlinelearning
Видео: vp6f, yuv420p, 500x273, 455 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[/spoiler]
[spoiler="PASCAL Bootcamp in Machine Learning, Marseille 2010"]
[spoiler="Introduction to Machine Learning by Iain Murray, 2010"]
Описание: How can we represent data on a computer and use it to learn to perform
useful tasks? This lecture reviews some simple classification and
regression rules, discusses under- and over-fitting and emphasises the
utility of defining objective functions for learning. There is also a
short overview of Bayesian learning, and some practical tips for
pre-processing and visualizing data. The lecture ends with a brief
mention of unsupervised learning and related topics.
Ссылка: http://videolectures.net/bootcamp2010_murray_iml
Видео: vp6f, yuv420p, 352x258, 455 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[/spoiler]
[spoiler="24th Annual Conference on Neural Information Processing Systems (NIPS), Vancouver 2010"]
[spoiler="Reinforcement Learning in Humans and Other Animals by Nathaniel Daw, 2011 (rec 2010)"]
Описание: Algorithms from computer science can serve as detailed process-level hypotheses for how the brain might approach difficult information processing problems. This tutorial reviews how ideas from the computational study of reinforcement learning have been used in biology to conceptualize the brain's mechanisms for trial-and-error decision making, drawing on evidence from neuroscience, psychology, and behavioral economics. We begin with the much-debated relationship between temporal-difference learning and the neuromodulator dopamine, and then consider how more sophisticated methods and concepts from RL -- including partial observability, hierarchical RL, function approximation, and various model-based approaches -- can provide frameworks for understanding additional issues in the biology of adaptive behavior.
In addition to helping to organize and conceptualize data from many different levels, computational models can be employed more quantitatively in the analysis of experimental data. The second aim of this tutorial is to review and demonstrate, again using the example of reinforcement learning, recent methodological advances in analyzing experimental data using computational models. An RL algorithm can be viewed as generative model for raw, trial-by-trial experimental data such as a subject's choices or a dopaminergic neuron's spiking responses; the problems of estimating model parameters or comparing candidate models then reduce to familiar problems in Bayesian inference. Viewed this way, the analysis of neuroscientific data is ripe for the application of many of the same sorts of inferential and machine learning techniques well studied by the NIPS community in other problem domains.
Ссылка: http://videolectures.net/nips2010_daw_rlh
Видео: vp6f, yuv420p, 352x258, 398 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[spoiler="Optimization Algorithms in Machine Learning by Stephen J. Wright, 2011 (rec 2010)"]
Описание: Optimization provides a valuable framework for thinking about,
formulating, and solving many problems in machine learning. Since
specialized techniques for the quadratic programming problem arising
in support vector classification were developed in the 1990s, there
has been more and more cross-fertilization between optimization and
machine learning, with the large size and computational demands of
machine learning applications driving much recent algorithmic research
in optimization. This tutorial reviews the major computational
paradigms in machine learning that are amenable to optimization
algorithms, then discusses the algorithmic tools that are being
brought to bear on such applications. We focus particularly on such
algorithmic tools of recent interest as stochastic and incremental
gradient methods, online optimization, augmented Lagrangian methods,
and the various tools that have been applied recently in sparse and
regularized optimization.
Ссылка: http://videolectures.net/nips2010_wright_oaml
Видео: vp6f, yuv420p, 352x264, 409 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[/spoiler]
[spoiler="CERN Summer School 2010"]
[spoiler="Introduction to Statistics by Glen Cowan, 2011 (rec 2010)"]
Ссылка: http://videolectures.net/cernstudentsummerschool2010_cowan_statistics
Видео: vp6f, yuv420p, 500x274, 455 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[/spoiler]
[spoiler="Machine Learning Summer School (MLSS), Bordeaux 2011"]
[spoiler="Graphical Models and message-passing algorithms by Martin J. Wainwright, 2011"]
Ссылка: http://videolectures.net/mlss2011_wainwright_messagepassing
Видео: vp6f, yuv420p, 500x280, 512 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[spoiler="Monte Carlo Methods by Arnaud Doucet, 2011"]
Описание: We will first review the Monte Carlo principle and standard Monte Carlo methods including rejection sampling, importance sampling and standard Markov chain Monte Carlo (MCMC) methods. We will then discuss more advanced MCMC methods such as adaptive MCMC methods and auxiliary variable methods such as parallel tempering, particle MCMC methods and slice sampling.
Ссылка: http://videolectures.net/mlss2011_doucet_montecarlo
Видео: vp6f, yuv420p, 500x280, 512 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[spoiler="Convex Optimization by Lieven Vandenberghe, 2011"]
Описание: The lectures will give an introduction to the theory and applications of convex optimization, and an overview of recent developments in algorithms.
The first lecture will cover the basics of convex analysis, focusing on the results that are most useful for convex modeling, i.e., recognizing and formulating convex optimization problems in applications. We will introduce conic optimization, and the two most widely studied types of conic optimization problems, second-order cone and semidefinite programs. The material will be illustrated with applications to robust optimization, convex relaxations in nonconvex optimization, and convex techniques for sparse optimization.
Lecture 2 will cover interior-point methods for conic optimization, including path-following methods and symmetric primal-dual methods, and the numerical implementation of interior-point methods.
Lecture 3 will focus on first-order algorithms for large-scale convex optimization, including recent developments in the area of proximal gradient methods, and on dual decomposition and multiplier methods.
Ссылка: http://videolectures.net/mlss2011_vandenberghe_convex
Видео: vp6f, yuv420p, 500x280, 512 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[spoiler="Learning Theory: statistical and game-theoretic approaches by Nicolò Cesa-Bianchi, 2011"]
Описание: The theoretical foundations of machine learning have a double nature: statistical and game-theoretic. In this course we take advantage of both paradigms to introduce and investigate a number of basic topics, including mistake bounds and risk bounds, empirical risk minimization, online linear optimization, compression bounds, overfitting and regularization.
The goal of the course is to provide a sound mathematical framework within which one can investigate basic questions in learning theory, such as the dependence of the predictive performance of a model on the complexity of the model class and on the amount of training information.
Ссылка: http://videolectures.net/mlss2011_cesa_bianchi_learningtheory
Видео: vp6f, yuv420p, 500x280, 512 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[spoiler="Kernel Methods by Bernhard Schölkopf, 2011"]
Описание: The course will start with basic ideas of machine learning, followed by some elements of learning theory. It will also introduce positive definite kernels and their associated feature spaces, and show how to use them for kernel mean embeddings, SVMs, and kernel PCA.
Ссылка: http://videolectures.net/mlss2011_scholkopf_kernel
Видео: vp6f, yuv420p, 500x280, 512 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[/spoiler]
[spoiler="NIPS Workshops, Sierra Nevada 2011"]
[spoiler="Optimization for Machine Learning"]
[spoiler="Making Gradient Descent Optimal for Strongly Convex Stochastic Optimization by Ohad Samir, 2012 (rec 2011)"]
Описание: Stochastic gradient descent (SGD) is a simple and popular method to solve
stochastic optimization problems which arise in machine learning. For strongly
convex problems, its convergence rate was known to be O(log(T)/T), by running
SGD for T iterations and returning the average point. However, recent results
showed that using a different algorithm, one can get an optimal O(1/T)
rate. This might lead one to believe that standard SGD is suboptimal, and maybe
should even be replaced as a method of choice. In this paper, we investigate the
optimality of SGD in a stochastic setting. We show that for smooth problems, the
algorithm attains the optimal O(1/T) rate. However, for non-smooth problems,
the convergence rate with averaging might really be
(log(T)/T), and this is not
just an artifact of the analysis. On the flip side, we show that a simple modification
of the averaging step suffices to recover the O(1/T) rate, and no other change of
the algorithm is necessary. We also present experimental results which support
our findings, and point out open problems.
Ссылка: http://videolectures.net/nipsworkshops2011_shamir_convex
Видео: vp6f, yuv420p, 500x280, 512 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[spoiler="Limited-memory quasi-Newton and Hessianfree Newton methods for non-smooth optimization by Mark Schmidt, 2011 (rec 2010)"]
Описание: Limited-memory quasi-Newton and Hessian-free Newton
methods are two workhorses of unconstrained optimization of
high-dimensional smooth objectives. However, in many cases
we would like to optimize a high-dimensional unconstrained
objective function that is non-smooth due to the presence
of a ‘simple’ non-smooth regularization term. Motivated by
problems arising in estimating sparse graphical models, in this
talk we focus on strategies for extending limited-memory quasi-
Newton and Hessian-free Newton methods for unconstrained
optimization to this scenario. We first consider two-metric (sub-)
gradient projection methods for problems where the regularizer is
separable, and then consider proximal Newton-like methods for
group-separable and non-separable regularizers. We will discuss
several applications where sparsity-encouraging regularizers are
used to estimate graphical model parameters and/or structure,
including the estimation of sparse, blockwise-sparse, and
structured-sparse models.
Ссылка: http://videolectures.net/nipsworkshops2010_schmidt_lmq
Видео: vp6f, yuv420p, 352x258, 398 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[spoiler="Lock-Free Approaches to Parallelizing Stochastic Gradient Descent by Benjamin Recht, 2012 (rec 2011)"]
Описание: Stochastic Gradient Descent (SGD) is a very popular optimization algorithm for solving data-driven machine learning problems. SGD is well suited to processing large amounts of data due to its robustness against noise, rapid convergence rates, and predictable memory footprint. Nevertheless, SGD seems to be impeded by many of the classical barriers to scalability: (1) SGD appears to be inherently sequential, (2) SGD assumes uniform sampling from the underlying data set resulting in poor locality, and (3) current approaches to parallelize SGD require performance-destroying, fine-grained communication.
This talk aims to refute the conventional wisdom that SGD inherently suffers from these impediments. Specifically, I will show that SGD can be implemented in parallel with minimal communication, with no locking or synchronization, and with strong spatial locality. I will provide both theoretical and experimental evidence demonstrating the achievement of linear speedups on multicore workstations on several benchmark optimization problems. Finally, I will close with a discussion of a challenging problem raised by our implementations relating arithmetic and geometric means of matrices.
Joint work with Feng Niu, Christopher Re, and Stephen Wright.
Ссылка: http://videolectures.net/nipsworkshops2011_recht_lockfree
Видео: vp6f, yuv420p, 500x280, 512 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[spoiler="Fast first-order methods for convex optimization with line search by Katya Scheinberg, 2012 (rec 2011)"]
Описание: We propose accelerated first-order methods with non-monotonic choice of the
prox parameter, which essentially controls the step size. This is in contrast with
most accelerated schemes where the prox parameter is either assumed to be constant
or non-increasing. In particular we show that a backtracking strategy can be
used within FISTA [2] and FALM algorithms [5] starting from an arbitrary parameter
value preserving their worst-case iteration complexities of O.
We also derive complexity estimates that depend on the “average” step size rather
than the global Lipschitz constant for the function gradient, which provide better
theoretical justification for these methods, hence the main contribution of this
paper is theoretical.
Ссылка: http://videolectures.net/nipsworkshops2011_scheinberg_convexoptimization
Видео: vp6f, yuv420p, 500x280, 512 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[spoiler="Efficiency of Quasi-Newton Methods on Strictly Positive Functions by Yurii Nesterov, 2011 (rec 2010)"]
Описание: In this talk we consider a new class of convex optimization
problems, which admit faster black-box optimization schemes.
For analyzing their rate of convergence, we introduce a notion of
mixed accuracy of an approximate solution, which is a convenient
generalization of the absolute and relative accuracies. We show
that for our problem class, a natural Quasi-Newton method is
always faster than the standard gradient method. At the same
time, after an appropriate normalization, our results can be
extended onto the general convex unconstrained minimization
problems.
Ссылка: http://videolectures.net/nipsworkshops2010_nesterov_eqn
Видео: vp6f, yuv420p, 352x258, 398 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[/spoiler]
[spoiler="Cosmology meets Machine Learning"]
[spoiler="Efficient Estimation of N-point Spatial Statistics by Alexander G. Gray, 2012 (rec 2011)"]
Описание: Precise statistical analyses of astronomical data are the key to validating models of complex phenomena, such as dark matter and dark energy. In particular, spatial statistics are needed for large-scale sky catalogs. The n-point correlation functions provide a complete description of any point process and are widely used to understand astronomical data. However, the computational cost of estimating these functions scales as N^n for N
data points. Furthermore, these expensive computations must be repeated many times at many different scales in order to gain a detailed picture of the correlation function and to estimate its variance. Since astronomy surveys contain hundreds of millions or billions of points (and are growing rapidly), these computations are infeasible. We present a new approach based on multidimensional trees to overcome these computational obstacles. We build on the previously most efficient algorithm (Gray and Moore, 2001, Moore, et al., 2001) which improved over the N^n scaling of a direct computation. In this work, we incorporate the computations at different scales along with the variance estimation directly. We can therefore achieve an order of magnitude speedup over the current state-of-the-art method. We show preliminary scaling results on a mock galaxy catalog.
Ссылка: http://videolectures.net/nipsworkshops2011_gray_efficient
Видео: vp6f, yuv420p, 500x280, 487 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[/spoiler]
[spoiler="Relations between machine learning problems – an approach to unify the field"]
[spoiler="Relations Betweeen Machine Learning Problems by Robert C. Williamson, 2012 (rec 2011)"]
Ссылка: http://videolectures.net/nipsworkshops2011_williamson_machine
Видео: vp6f, yuv420p, 500x280, 487 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 131 kb/s
[/spoiler]
[/spoiler]
[/spoiler]
[spoiler="Machine Learning seminars at the Cambridge University Engineering Department"]
[spoiler="Group Theory and Machine Learning by Risi Kondor, 2008 (rec 2007)"]
Описание: Machine Learning Tutorial Lecture
The use of algebraic methods—specifically group theory, representation theory, and even some concepts from algebraic geometry—is an emerging new direction in machine learning. The purpose of this tutorial is to give an entertaining but informative introduction to the background to these developments and sketch some of the many possible applications, including multi-object tracking, learning rankings, and constructing translation and rotation invariant features for image recognition. The tutorial is intended to be palatable by a non-specialist audience with no prior background in abstract algebra.
Ссылка: http://videolectures.net/mlcued08_kondor_gtm
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="MIT World Host: Computer Science and Artificial Intelligence Laboratory (CSAIL)"]
[spoiler="Emotion Machine: Commonsense Thinking, Artificial Intelligence, and the Future of the Human Mind by Marvin Minsky, 2011 (rec 2007)"]
Описание: Contemporary artificial intelligence researchers (as well as neurologists and Karl Jung) are taken to task in this talk by one of the world’s preeminent scholars of artificial intelligence.
Marvin Minsky is worried that after making great strides in its infancy, AI has lost its way, getting bogged down in different theories of machine learning. Researchers “have tried to invent single techniques that could deal with all problems, but each method works only in certain domains.” Minsky believes we’re facing an AI emergency, since soon there won’t be enough human workers to perform the necessary tasks for our rapidly aging population.
So while we have a computer program that can beat a world chess champion, we don’t have one that can reach for an umbrella on a rainy day, or put a pillow in a pillow case. For “a machine to have common sense, it must know 50 million such things,” and like a human, activate different kinds of expertise in different realms of thought, says Minsky.
Minsky suggests that such a machine should, like humans, have a very high-level, rule-based system for recognizing certain kinds of problems. He labels these parts of the brain “critics.” When one critic gets selected in a particular situation, the others get turned off. In the “cloud of resources” that comprises our mind, mental states, from emotions to reasoning, result from activating or suppressing the right resource. Minsky further refines his machine’s reasoning architecture with six levels of thinking that attempt to emulate the different kinds of reasoning humans may engage in, often simultaneously: These include learned reactions, deliberative thinking, and reflective thinking, among others. A smart machine must have at least these levels, he says, because psychology, unlike physics, doesn’t lend itself to a minimal number of laws. With at least 400 different areas of the brain operating, “if a theory tries to explain everything by just 20 principles, it’s doing something wrong.”
Today, while we have machines that can automatically assemble clothes, we don’t have any that know how to sew together a tear in a shirt or a suit. Minsky proposes a new kind of AI that might eventually result in a “really resourceful, clever thinking machine...with knowledge about how to do things,” and which “can do the broad range of things children can do.”
Ссылка: http://videolectures.net/mitworld_minsky_emoticon
Видео: vp6f, yuv420p, 480x360, 368 kb/s, 15 tbr, 1k tbn, 1k tbc
Аудио: mp3, 22050 Hz, stereo, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="Single Lectures Series"]
[spoiler="A tutorial on Deep Learning by Geoffrey E. Hinton, 2009"]
Описание: Complex probabilistic models of unlabeled data can be created by combining simpler models. Mixture models are obtained by averaging the densities of simpler models and "products of experts" are obtained by multiplying the densities together and renormalizing. A far more powerful type of combination is to form a "composition of experts" by treating the values of the latent variables of one model as the data for learning the next model. The first half of the tutorial will show how deep belief nets -- directed generative models with many layers of hidden variables -- can be learned one layer at a time by composing simple, undirected, product of expert models that only have one hidden layer. It will also explain why composing directed models does not work. Deep belief nets are trained as generative models on large, unlabeled datasets, but once multiple layers of features have been created by unsupervised learning, they can be fine-tuned to give excellent discrimination on small, labeled datasets. The second half of the tutorial will describe applications of deep belief nets to several tasks including object recognition, non-linear dimensionality reduction, document retrieval, and the interpretation of medical images. It will also show how the learning procedure for deep belief nets can be extended to high-dimensional time series and hierarchies of Conditional Random Fields.
Ссылка: http://videolectures.net/jul09_hinton_deeplearn
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Learning Deep Hierarchies of Representations by Yoshua Bengio, 2009"]
Описание: Whereas theoretical work suggests that deep architectures might be computationally and statistically more efficient at representing highly-varying functions, training deep architectures was unsuccessful until the recent advent of algorithms based on unsupervised pre-training of each level of a hierarchically structured model. Several unsupervised criteria and procedures were proposed for this purpose, starting with the Restricted Boltzmann Machine (RBM), which when stacked gives rise to Deep Belief Networks (DBN). Although the partition function of RBMs is intractable, inference is tractable and we review several successful learning algorithms that have been proposed, in particular those using weights that change quickly during learning instead of converging. In addition to being impressive as generative models, deep architectures based on RBMs and other unsupervised learning methods have made an impact by being used to initialize deep supervised neural networks. Even though these new algorithms have enabled training deep models, many questions remain as to the nature of this difficult learning problem. We attempt to shed some light on these questions by comparing different successful approaches to training deep architectures and through extensive simulations investigating explanatory hypotheses. Finally, we describe our current research program, objectives and challenges, regarding learning representations at multiple levels of abstraction, to compare web objects such as images, documents, and search engine requests, comparisons that are at the core of several information retrieval applications.
Ссылка: http://videolectures.net/okt09_bengio_ldhr
Видео: h264 (High), yuv420p, 480x330 [PAR 1:1 DAR 16:11], 200 kb/s, 29.97 tbr, 1k tbn, 59.94 tbc
Аудио: mp3, 22050 Hz, stereo, s16, 64 kb/s
[/spoiler]
[/spoiler]
[spoiler="NATO Advanced Study Institute on Mining Massive Data Sets for Security"]
[spoiler="Foundations of Statistical Learning Theory - Empirical Inference in high-dimention spaces by Léon Bottou, Vladimir Vapnik, 2007"]
Ссылка: http://videolectures.net/mmdss07_bottou_fslt
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Learning using Many Examples by Léon Bottou, 2007"]
Описание: The statistical learning theory suggests to choose large capacity
models that barely avoid over-fitting the training data. In that perspective,
all datasets are small. Things become more complicated when
one considers the computational cost of processing large datasets.
Computationally challenging training sets appear when one want to emulate
intelligence: biological brains learn quite efficiently from the continuous
streams of perceptual data generated by our six senses, using
limited amounts of sugar as a source of power. Computationally challenging
training sets also appear when one want to analyze the masses
of data that describe the life of our computerized society. The more data
we understand, the more we enjoy competitive advantages.
– The first part of the tutorial clarifies the relation between the statistical
efficiency, the design of learning algorithms and their computational
cost.
– The second part makes a detailed exploration of specific learning algorithms
and of their implementation, with both simple and complex
examples.
– The third part considers algorithms that learn with a single pass
over the data. Certain algorithms have optimal properties but are
often too costly. Workarounds are discussed.
– Finally, the fourth part shows how active example selection provides
greater speed and reduces the feedback pressure that constrain parallel
implementations.
Ссылка: http://videolectures.net/mmdss07_bottou_lume
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[/spoiler]
[spoiler="The 13th International Conference on Knowledge Discovery and Data Mining"]
[spoiler="From Trees to Forests and Rule Sets - A Unified Overview of Ensemble Methods by John Elder, Giovanni Seni, 2007"]
Описание: Ensemble methods are one of the most influential developments in Machine Learning over the past decade. They perform extremely well in a variety of problem domains, have desirable statistical properties, and scale well computationally. By combining competing models into a committee, they can strengthen “weak” learning procedures.
This tutorial is aimed at both novice and advanced data mining researchers and practitioners especially in Engineering, Statistics, and Computer Science. Users with little exposure to ensemble methods will gain a clear overview of each method. Advanced practitioners already employing ensembles will gain insight into this breakthrough way to create next-generation models.
Ссылка: http://videolectures.net/kdd07_elder_seni_fttf
Видео: vp6f, yuv420p, 352x288, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, mono, s16, 65 kb/s
[/spoiler]
[spoiler="Learning Bayesian Networks by Richard E. Neapolitan, 2007"]
Описание: Bayesian networks are graphical structures for representing the probabilistic relationships among a large number of variables and doing probabilistic inference with those variables. The 1990's saw the emergence of excellent algorithms for learning Bayesian networks from passive data.
I will discuss the constraint-based learning method using an intuitive approach that concentrates on causal learning. Then I will discuss the Bayesian approach with some simple examples. I will show how, using the Bayesian approach, we can even learning something about causal influences from passive data on two variables. Finally, I will show some applications to finance and marketing.
Ссылка: http://videolectures.net/kdd07_neapolitan_lbn
Видео: vp6f, yuv420p, 320x256, 409 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[/spoiler]
[spoiler="Autumn School 2006: Machine Learning over Text and Images - Pittsburgh"]
[spoiler="Semisupervised Learning Approaches by Tom Mitchell, 2007 (rec 2006)"]
Ссылка: http://videolectures.net/mlas06_mitchell_sla
Видео: vp6f, yuv420p, 320x240, 307 kb/s, 25 tbr, 1k tbn, 1k tbc
Аудио: mp3, 44100 Hz, stereo, s16, 98 kb/s
[/spoiler]
[/spoiler]
[/spoiler]
[spoiler="Скриншоты"]






[/spoiler]
Доп. информация: Если из название выступления всё ещё не ясно о чём же видео и нету описания то остаётся ещё одна возможность: в папке с видео скорее всего есть файл slides_descr.txt с заголовками слайдов.

Пока скачано то, что показалось мне интересным. Если есть желание выложить какое-нибудь другое видео с videolectures.net - пишите в ЛС или сюда. Таким образом раздача возмножно будет обновляться в будущем

UPD: 26.10.2012 обновлены описания