In many applications, the data, such as web pages and research papers, contain relation (link) structure among entities in addition to textual content information. Matrix factorization (MF) methods, such as latent semantic indexing (LSI), have been successfully used to map either content information or relation information into a lower-dimensional latent space for subsequent processing. However, how to simultaneously model both the relation information and the content information effectively with an MF framework is still an open research problem. In this paper, we propose a novel MF method called relation regularized matrix factorization (RRMF) for relational data analysis. By using relation information to regularize the content MF procedure, RRMF seamlessly integrates both the relation information and the content information into a principled framework. We propose a linear-time learning algorithm with convergence guarantee to learn the parameters of RRMF. Extensive experiments on real data sets show that RRMF can achieve state-of-the-art performance.

In many real-world situations we are charged with detecting change as soon as possible. Important examples include detecting medical conditions, detecting security breaches, and updating caches of distributed databases. In those situations, sensing can be expensive, but it is also important to detect change in a timely manner. In this paper we present tractable greedy algorithms and prove that they solve this decision problem either optimally or approximate the optimal solution in many cases. Our problem model is a POMDP that includes a cost for sensing, a cost for delayed detection, a reward for successful detection, and no-cost partial observations. Making optimal decisions is difficult in general. We show that our tractable greedy approach finds optimal policies for sensing both a single variable and multiple correlated variables. Further, we provide approximations for the optimal solution to multiple hidden or observed variables per step. Our algorithms outperform previous algorithms in experiments over simulated data and live Wikipedia WWW pages.

The Possible Winner problem asks whether some distinguished candidate may  become the winner of an election when the given incomplete votes are extended into complete ones in a favorable way. Possible Winner is NP-complete for common voting rules such as Borda, many other positional scoring rules, Bucklin, Copeland etc.  We investigate how three different parameterizations influence the computational complexity of Possible Winner for a number of voting rules.  We show  fixed-parameter tractability results with respect to the parameter "number of candidates" but intractability results with respect to the parameter "number of votes". Finally, we derive fixed-parameter tractability results with respect to the parameter "total number of undetermined candidate pairs" and identify an interesting polynomial-time solvable special case for Borda.

This paper deals with  multiobjective optimization in the context of multiattribute utility theory. The alternatives (feasible solutions) are seen as elements of a product set of attributes and preferences over solutions are represented by generalized additive decomposable (GAI) utility functions modeling individual preferences or criteria. Due to decomposability, utility vectors attached to solutions can be compiled into a graphical structure closely related to junction trees, the so-called GAI net. We first show how the structure of the GAI net can be used to determine efficiently the exact set of Pareto-optimal solutions in a product set and provide numerical tests on random instances. Since the exact determination of the Pareto set is intractable in worst case, we propose a near admissible algorithm with performance guarantee, exploiting the GAI structure to approximate the set of Pareto optimal solutions. We present numerical experimentations, showing that both utility decomposition and approximation significantly improve resolution times in multiobjective search problems.

The Universum sample, which is defined as the sample that doesn't belong to any of the classes the learning task concerns, has been proved to be helpful in both supervised and semi-supervised settings. The former works treat the Universum samples equally. Our research found that not all the Universum samples are helpful, and we propose a method to pick the informative ones, i.e., in-between Universum samples. We also set up a new semi-supervised framework to incorporate the in-between Universum samples. Empirical experiments show that our method outperforms the former ones.

Evaluating an agent's performance in a stochastic setting is necessary for agent development, scientific evaluation, and competitions. Traditionally, evaluation is done using Monte Carlo estimation; the magnitude of the stochasticity in the domain or the high cost of sampling, however, can often prevent the approach from resulting in statistically significant conclusions. Recently, an advantage sum technique has been proposed for constructing unbiased, low variance estimates of agent performance. The technique requires an expert to define a value function over states of the system, essentially a guess of the state's unknown value. In this work, we propose learning this value function from past interactions between agents in some target population. Our learned value functions have two key advantages: they can be applied in domains where no expert value function is available and they can result in tuned evaluation for a specific population of agents (e.g., novice versus advanced agents). We demonstrate these two advantages in the domain of poker. We show that we can reduce variance over state-of-the-art estimators for a specific population of limit poker players as well as construct the first variance reducing estimators for no-limit poker and multi-player limit poker.

In this paper we approach the problem of reasoning with quantified Boolean formulas (QBFs) by combining search and resolution, and by switching between them according to structural properties of QBFs. We provide empirical evidence that QBFs which cannot be solved by search or resolution alone, can be solved by combining them, and that our approach makes a proof-of-concept implementation competitive with current QBF solvers.

Recently, Argumentation Mechanism Design (ArgMD) was introduced as a new paradigm for studying argumentation among self-interested agents using game-theoretic techniques. Preliminary results showed a condition under which a direct mechanism based on Dung's grounded semantics is strategy-proof (i.e. truth enforcing). But these early results dealt with a highly restricted form of agent preferences, and assumed agents can only hide, but not lie about, arguments. In this paper, we characterise strategy-proofness under grounded semantics for a more realistic preference class (namely, focal arguments). We also provide the first analysis of the case where agents can lie.

In many real world planning domains, some observation information is optional and useless to the execution of a plan; on the other hand, information acquisition may require some kind of cost. The problem of observation reduction for strong plans has been addressed in the literature. However, observation reduction for plans with contexts (which are more general and useful than strong plans in robotics) is still a open problem. In this paper, we present an attempt to solve the problem. Our first contribution is the definition of structured plans, which can encode sequential, conditional and iterative behaviors, and is expressive enough for dealing with incomplete observation information and internal states of the agent. A second contribution is an observation reduction algorithm for plans with contexts, which can transform a plan with contexts into a structured plan that only branches on necessary observation information.

Gaussian Processes (GPs) are promising Bayesian methods for classification and regression problems. They have also been used for semi-supervised learning tasks. In this paper, we propose a new algorithm for solving semi-supervised binary classification problem using sparse GP regression (GPR) models. It is closely related to semi-supervised learning based on support vector regression (SVR) and maximum margin clustering. The proposed algorithm is simple and easy to implement. It gives a sparse solution directly unlike the SVR based algorithm. Also, the hyperparameters are estimated easily without resorting to expensive cross-validation technique. Use of sparse GPR model helps in making the proposed algorithm scalable. Preliminary results on synthetic and real-world data sets demonstrate the efficacy of the new algorithm.