Reinforcement learning addresses the problem of learning to select actions in order to maximize one’s performance in unknown environments. To scale reinforcement learning to complex real-world tasks, such as typically studied in AI, one must ultimately be able to discover the structure in the world, in order to abstract away the myriad of details and to operate in more tractable problem spaces....

We investigated the extent to which the ability to perceive spatial form from temporal structure (TS) improves with practice. Observers trained monocularly for a number of consecutive days on a shape discrimination task, with one group of observers judging shape defined by luminance contrast between target and background elements and another group judging shape defined by correlated TS (synchro...

To learn a visual code in an unsupervised manner, one may attempt to capture those features of the stimulus set that would contribute significantly to a statistically efficient representation (as dictated, e.g., by the Minimum Description Length principle). Paradoxically, all the candidate features in this approach need to be known before statistics over them can be computed. This paradox may b...

We present a robust multiple manifolds structure learning (RMMSL) scheme to robustly estimate data structures under the multiple low intrinsic dimensional manifolds assumption. In the local learning stage, RMMSL efficiently estimates local tangent space by weighted low-rank matrix factorization. In the global learning stage, we propose a robust manifold clustering method based on local structur...

In this paper, we empirically evaluate effectiveness of structure learning of Bayesian Network when applying such networks to the domain of Keystroke Dynamics authentication. We compare four structure learning methods of Bayesian Network Classifier – Genetic, TAN, K2, and Hill Climbing algorithms, on our authentication model, namely Classify User via Short-text and IP Model (CUSIM). The results...

We present Minimum Description Length techniques for learning the structure of weighted languages. MDL is already widely used both for segmentation and classification tasks, and here we show it can be used to formalize further important tools in the descriptive linguists’ toolbox, including the distinction between accidental and systematic gaps in the data, the detection of ambiguity, the selec...

Structure Learning in Sequential Data Liam Stewart Master of Science Graduate Department of Computer Science University of Toronto 2005 The goal of discriminative sequence learning is to learn how to classify items that can be arranged in a sequence. Many models have been proposed including logistic regression, the maximum entropy Markov model, the conditional random field, the input output Mar...

Graph-based representations have been used with considerable success in computer vision in the abstraction and recognition of object shape and scene structure. Despite this, the methodology available for learning structural representations from sets of training examples is relatively limited. In this paper we take a simple yet effective Bayesian approach to attributed graph learning. We present...

We study the structure learning problem for graph homomorphisms, commonly referred to as Hcolorings, including the weighted case which corresponds to spin systems with hard constraints. The learning problem is as follows: for a fixed (and known) constraint graphH with q colors and an unknown graph G = (V,E) with n vertices, given uniformly random H-colorings of G, how many samples are required ...

Discriminative Machine Learning with Structure by Simon Lacoste-Julien Doctor of Philosophy in Computer Science and the Designated Emphasis in Communication, Computation and Statistics University of California, Berkeley Professor Michael I. Jordan, Chair Some of the best performing classifiers in modern machine learning have been designed using discriminative learning, as exemplified by Support...