The overall goal of our research is the automatic recognition of the intended message of a grouped bar chart. This paper presents our preliminary work on a system that utilizes the communicative signals in a grouped bar chart as evidence in a Bayesian network that hypothesizes the primary message conveyed by the graphic. The paper discusses the kinds of communicative signals present in grouped bar charts and an ACT-R model for computationalizing one important communicative signal, the relative effort involved in performing the perceptual tasks necessary for the recognition. It also describes our Bayesian network and its implementation on a subset of the kinds of messages that can be conveyed by grouped bar charts.

Exploiting ontologies for efficient inference is one of the most widely studied topics in knowledge representation and reasoning. The use of ontologies for probabilistic inference, however, is much less developed. A number of algorithms for lifted inference in first-order probabilistic languages have been proposed, but their scalability is limited by the combinatorial explosion in the sets of objects that need to be considered. We propose a coarse-to-fine inference approach that leverages a class hierarchy to combat this problem. Starting at the highest level, our approach performs inference at successively finer grains, pruning low-probability atoms before refining. We provide bounds on the error incurred by this approach relative to full ground inference as a function of the pruning threshold. We also show how to learn parameters in a coarse-to-fine manner to maximize the opportunities for pruning during inference. Experiments on link prediction and biomolecular event prediction tasks show our method can greatly improve the scalability of lifted probabilistic inference.

Knowledge compilation is a valuable tool for dealing with the computational intractability of propositional reasoning. In knowledge compilation, a representation in a source language is typically compiled into a target language in order to perform some reasoning task in polynomial time. One particularly popular target language is Deterministic Decomposable Negation Normal Form (d-DNNF). d-DNNF supports efficient reasoning for tasks such as consistency checking and model counting, and as such it has proven a useful representation language for Bayesian inference, conformant planning, and diagnosis. In this paper, we exploit recent advances in #SAT solving in order to produce a new state-of-the-art CNF → d-DNNF compiler. We evaluate the properties and performance of our compiler relative to C2D, the de facto standard for compiling to d-DNNF. Empirical results demonstrate that our compiler is generally one order of magnitude faster than C2D on typical benchmark problems while yielding a d-DNNF representation of comparable size.

The projects that we have developed for UC Berkeley’s introductory artificial intelligence (AI) course teach foundational concepts using the classic video game Pac-Man. There are four project topics: state-space search, multi-agent search, probabilistic inference, and reinforcement learning. Each project requires students to implement general-purpose AI algorithms and then to inject domain knowledge about the Pac- Man environment using search heuristics, evaluation functions, and feature functions. We have found that the Pac-Man theme adds consistency to the course, as well as tapping in to students’ excitement about video games.

As network attacks have increased in number and severity over the past few years, intrusion detection system (IDS) is increasingly becoming a critical component to secure the network. Due to large volumes of security audit data as well as complex and dynamic properties of intrusion behaviors, optimizing performance of IDS becomes an important open problem that is receiving more and more attention from the research community. The uncertainty to explore if certain algorithms perform better for certain attack classes constitutes the motivation for the reported herein. In this paper, we evaluate performance of a comprehensive set of classifier algorithms using KDD99 dataset. Based on evaluation results, best algorithms for each attack category is chosen and two classifier algorithm selection models are proposed. The simulation result comparison indicates that noticeable performance improvement and real-time intrusion detection can be achieved as we apply the proposed models to detect different kinds of network attacks.

Classic decision-theory is based on the maximum expected utility (MEU) principle, but crucially ignores the resource costs incurred when determining optimal decisions. Here we propose an axiomatic framework for bounded decision-making that considers resource costs. Agents are formalized as probability measures over input-output streams. We postulate that any such probability measure can be assigned a corresponding conjugate utility function based on three axioms: utilities should be real-valued, additive and monotonic mappings of probabilities. We show that these axioms enforce a unique conversion law between utility and probability (and thereby, information). Moreover, we show that this relation can be characterized as a variational principle: given a utility function, its conjugate probability measure maximizes a free utility functional. Transformations of probability measures can then be formalized as a change in free utility due to the addition of new constraints expressed by a target utility function. Accordingly, one obtains a criterion to choose a probability measure that trades off the maximization of a target utility function and the cost of the deviation from a reference distribution. We show that optimal control, adaptive estimation and adaptive control problems can be solved this way in a resource-efficient way. When resource costs are ignored, the MEU principle is recovered. Our formalization might thus provide a principled approach to bounded rationality that establishes a close link to information theory.

The book deals with the supervised-learning problem for both regression and classification, and includes detailed algorithms. A wide variety of covariance (kernel) functions are presented and their properties discussed. Model selection is discussed both from a Bayesian and a classical perspective. Many connections to other well-known techniques from machine learning and statistics are discussed, including support-vector machines, neural networks, splines, regularization networks, relevance vector machines and others. Theoretical issues including learning curves and the PAC-Bayesian framework are treated, and several approximation methods for learning with large datasets are discussed.

In this paper we treat both forms of probabilistic inference, estimating marginal probabilities of the joint distribution and finding the most probable assignment, through a unified message-passing algorithm architecture. We generalize the Belief Propagation (BP) algorithms of sum-product and max-product and tree-rewaighted (TRW) sum and max product algorithms (TRBP) and introduce a new set of convergent algorithms based on "convex-free-energy" and Linear-Programming (LP) relaxation as a zero-temprature of a convex-free-energy. The main idea of this work arises from taking a general perspective on the existing BP and TRBP algorithms while observing that they all are reductions from the basic optimization formula of $f + \sum_i h_i$ where the function $f$ is an extended-valued, strictly convex but non-smooth and the functions $h_i$ are extended-valued functions (not necessarily convex). We use tools from convex duality to present the "primal-dual ascent" algorithm which is an extension of the Bregman successive projection scheme and is designed to handle optimization of the general type $f + \sum_i h_i$. Mapping the fractional-free-energy variational principle to this framework introduces the "norm-product" message-passing. Special cases include sum-product and max-product (BP algorithms) and the TRBP algorithms. When the fractional-free-energy is set to be convex (convex-free-energy) the norm-product is globally convergent for estimating of marginal probabilities and for approximating the LP-relaxation. We also introduce another branch of the norm-product, the "convex-max-product". The convex-max-product is convergent (unlike max-product) and aims at solving the LP-relaxation.

We tackle the problem of multi-class relational sequence learning using relevant patterns discovered from a set of labelled sequences. To deal with this problem, firstly each relational sequence is mapped into a feature vector using the result of a feature construction method. Since, the efficacy of sequence learning algorithms strongly depends on the features used to represent the sequences, the second step is to find an optimal subset of the constructed features leading to high classification accuracy. This feature selection task has been solved adopting a wrapper approach that uses a stochastic local search algorithm embedding a naive Bayes classifier. The performance of the proposed method applied to a real-world dataset shows an improvement when compared to other established methods, such as hidden Markov models, Fisher kernels and conditional random fields for relational sequences.

The major challenge in designing a discriminative learning algorithm for predicting structured data is to address the computational issues arising from the exponential size of the output space. Existing algorithms make different assumptions to ensure efficient, polynomial time estimation of model parameters. For several combinatorial structures, including cycles, partially ordered sets, permutations and other graph classes, these assumptions do not hold. In this thesis, we address the problem of designing learning algorithms for predicting combinatorial structures by introducing two new assumptions: (i) The first assumption is that a particular counting problem can be solved efficiently. The consequence is a generalisation of the classical ridge regression for structured prediction. (ii) The second assumption is that a particular sampling problem can be solved efficiently. The consequence is a new technique for designing and analysing probabilistic structured prediction models. These results can be applied to solve several complex learning problems including but not limited to multi-label classification, multi-category hierarchical classification, and label ranking.