From: VideoLectures on Sat, Jan 22 2011 12:18 PM

From: VideoLectures on Sat, Jan 22 2011 12:17 PM

European archivists grapple with the legal obligations, civic responsibilities and future prospects of their collections, which, thanks to the Internet and other new technologies, are increasingly awash in image and sound. As William Urichhio notes, “tradition-bound institutions know what we sho...

From: VideoLectures on Sat, Jan 15 2011 3:05 AM

Automatic speech recognition has gradually improved over the years, but the reliable recognition of unconstrained speech is still not within reach. In order to achieve a breakthrough, many research groups are now investigating new methodologies that have potential to outperform the Hidden Markov...

From: VideoLectures on Sat, Jan 15 2011 3:04 AM

Relative to the large literature on upper bounds on complexity of convex optimization, lesser attention has been paid to the fundamental hardness of these problems. Recent years have seen a surge in optimization methods tailored to sparse optimization problems. In this paper, we study the comple...

From: VideoLectures on Sat, Jan 15 2011 3:04 AM

Principal Component Analysis is one of the most widely used techniques for dimensionality reduction. Nevertheless, it is plagued by sensitivity to outliers; finding robust analogs, particularly for high-dimensional data, is critical. We discuss the challenges posed by the high dimensional settin...

From: VideoLectures on Sat, Jan 15 2011 3:04 AM

Distributing labeling tasks among hundreds or thousands of annotators is an increasingly important method for annotating large datasets. We present a method for estimating the underlying value (e.g. the class) of each image from (noisy) annotations provided by multiple annotators. Our method is b...

From: VideoLectures on Sat, Jan 15 2011 3:04 AM

The Conformal Predictions framework is a recent development in machine learning to associate reliable measures of confidence with results in classification and regression. This framework is founded on the principles of algorithmic randomness (Kolmogorov complexity), transductive inference and hy...

From: VideoLectures on Sat, Jan 15 2011 3:04 AM

From: VideoLectures on Sat, Jan 15 2011 3:03 AM

From: VideoLectures on Sat, Jan 15 2011 3:03 AM

We consider the design of Markov chain Monte Carlo (MCMC) methods for large-scale, distributed, heterogeneous compute facilities, with a focus on synthesising sample sets across multiple runs performed in parallel. While theory suggests that many independent Markov chains may be run and their sa...

From: VideoLectures on Sat, Jan 15 2011 3:03 AM

From: VideoLectures on Sat, Jan 15 2011 3:03 AM

Recent work in reinforcement learning has emphasized the power of L1 regularization to perform feature selection and prevent overfitting. We propose formulating the L1 regularized linear fixed point problem as a linear complementarity problem (LCP). This formulation offers several advantages over...

From: VideoLectures on Sat, Jan 15 2011 3:03 AM

For large data it can be very time consuming to run gradient based optimizat ion,for example to minimize the log-likelihood for maximum entropy models.Distributed methods are therefore appealing and a number of distributed gradientoptimization strategies have been proposed including: distributed ...

From: VideoLectures on Sat, Jan 15 2011 3:02 AM

Professor Saltzman begins the lecture with discussion of the importance of motion for the survival and propagation of any living species. He presents the different modes of motion, taking first the example flight to talk about force balance, such as the magnitude of propulsive force that must be...

From: VideoLectures on Fri, Jan 14 2011 3:07 AM

Blur in photos due to camera shake, blur in astronomical image sequences due to atmospheric turbulence, and blur in magnetic resonance imaging sequences due to object motion are examples of blur that can not be adequately described as a space-invariant convolution, because such blur varies acros...

From: VideoLectures on Fri, Jan 14 2011 3:06 AM

From: VideoLectures on Fri, Jan 14 2011 3:05 AM

We consider the multiple linear regression problem, in a setting where some of the set of relevant features could be shared across the tasks. A lot of recent research has studied the use of L1 Lq norm block-regularizations with q and 1 for such (possibly) block-structured problems, establishing ...

From: VideoLectures on Fri, Jan 14 2011 3:05 AM

Very large informatics graphs such as large social and information networks typically have properties that render many popular machine learning and data analysis tools largely inappropriate. While this is problematic for these applications, it also suggests that these graphs may be useful as a t...

From: VideoLectures on Fri, Jan 14 2011 3:05 AM

In distributed averaging and consensus algorithms, processors exchange and update certain values (or "estimates", or "opinions") by forming a local average with the values of their neighbors. Under suitable conditions, such algorithms converge to consensus (every processor ends up holding the sa...

From: VideoLectures on Fri, Jan 14 2011 3:05 AM

We present an incremental subgradient algorithm for approximate computation of maximum-a-posteriori (MAP) states in cyclic graphical models. Its most striking property is its immense simplicity: each iteration requires only the solution of a sequence of trivial optimization problems. The algorith...