Ensemble methods for supervised machine learning have become popular due to their ability to accurately predict class labels with groups of simple, lightweight "base learners." While ensembles offer computationally efficient models that have good predictive capability they tend to be large and offer little insight into the patterns or structure in a dataset. We consider an ensemble technique that returns a model of ranked rules. The model accurately predicts class labels and has the advantage of indicating which parameter constraints are most useful for predicting those labels. An example of the rule ensemble method successfully ranking rules and selecting attributes is given with a dataset containing images of potential supernovas where the number of necessary features is reduced from 39 to 21. We also compare the rule ensemble method on a set of multi-class problems with boosting and bagging, which are two well known ensemble techniques that use decision trees as base learners, but do not have a rule ranking scheme.

Publishing private data on external servers incurs the problem of how to avoid unwanted disclosure of confidential data. We study a problem of confidentiality in extended disjunctive logic programs and show how it can be solved by extended abduction. In particular, we analyze how credulous non-monotonic reasoning affects confidentiality.

In order to give appropriate semantics to qualitative conditionals of the form "if A then normally B", ordinal conditional functions (OCFs) ranking the possible worlds according to their degree of plausibility can be used. An OCF accepting all conditionals of a knowledge base R can be characterized as the solution of a constraint satisfaction problem. We present a high-level, declarative approach using constraint logic programming techniques for solving this constraint satisfaction problem. In particular, the approach developed here supports the generation of all minimal solutions; these minimal solutions are of special interest as they provide a basis for model-based inference from R.

Statistical relational learning techniques have been successfully applied in a wide range of relational domains. In most of these applications, the human designers capitalized on their background knowledge by following a trial-and-error trajectory, where relational features are manually defined by a human engineer, parameters are learned for those features on the training data, the resulting model is validated, and the cycle repeats as the engineer adjusts the set of features. This paper seeks to streamline application development in large relational domains by introducing a light-weight approach that efficiently evaluates relational features on pieces of the relational graph that are streamed to it one at a time. We evaluate our approach on two social media tasks and demonstrate that it leads to more accurate models that are learned faster.

In this paper we present a proposal for a knowledge-based programming environment. In such an environment, declarative background knowledge, procedures, and concrete data are represented in suitable languages and combined in a flexible manner. This leads to a highly declarative programming style. We illustrate our approach on an example and report about our prototype implementation.

Combinatory categorial grammar (CCG) is a grammar formalism used for natural language parsing. CCG assigns structured lexical categories to words and uses a small set of combinatory rules to combine these categories to parse a sentence. In this work we propose and implement a new approach to CCG parsing that relies on a prominent knowledge representation formalism, answer set programming (ASP) - a declarative programming paradigm. We formulate the task of CCG parsing as a planning problem and use an ASP computational tool to compute solutions that correspond to valid parses. Compared to other approaches, there is no need to implement a specific parsing algorithm using such a declarative method. Our approach aims at producing all semantically distinct parse trees for a given sentence. From this goal, normalization and efficiency issues arise, and we deal with them by combining and extending existing strategies. We have implemented a CCG parsing tool kit - AspCcgTk - that uses ASP as its main computational means. The C&C supertagger can be used as a preprocessor within AspCcgTk, which allows us to achieve wide-coverage natural language parsing.

The problem has applications ranging from modeling pairwise comparisons to reconstructing mental processing architectures to conjoint testing. A necessary and sufficient condition for a given pattern of selective influences is provided by the Joint Distribution Criterion, according to which the problem of "what influences what" is equivalent to that of the existence of a joint distribution for a certain set of random variables. For inputs and outputs with finite sets of values this criterion translates into a test of consistency of a certain system of linear equations and inequalities (Linear Feasibility Test) which can be performed by means of linear programming. The Joint Distribution Criterion also leads to a metatheoretical principle for generating a broad class of necessary conditions (tests) for diagrams of selective influences. Among them is the class of distance-type tests based on the observation that certain functionals on jointly distributed random variables satisfy triangle inequality. A B C The Greek letters in this diagram represent inputs, or external factors, e.g., parameters of stimuli whose values can be chosen at will, or randomly vary but can be observed. The capital Roman letters stand for random outputs characterizing reactions of the system (an observer, a group of observers, a technical device, etc.). The arrows show which factor influences which random output. The factors are treated as deterministic entities: even if α,β,γ,δ in reality vary randomly (e.g., being randomly generated by a computer program, or being concomitant parameters of observations, such as age of respondents), for the purposes of analyzing selective influences the random outputs A, B,C are always viewed as conditioned upon various combinations of specific values of α,β,γ,δ. The first question to ask is: what is the meaning of the above diagram if the random outputs A,B,C in it are not necessarily stochastically independent?

This article presents a survey of work on lifted graphical models. We review a general form for a lifted graphical model, a par-factor graph, and show how a number of existing statistical relational representations map to this formalism. We discuss inference algorithms, including lifted inference algorithms, that efficiently compute the answers to probabilistic queries. We also review work in learning lifted graphical models from data. It is our belief that the need for statistical relational models (whether it goes by that name or another) will grow in the coming decades, as we are inundated with data which is a mix of structured and unstructured, with entities and relations extracted in a noisy manner from text, and with the need to reason effectively with this data. We hope that this synthesis of ideas from many different research groups will provide an accessible starting point for new researchers in this expanding field.

Comparative Evaluation of Prediction Algorithms(COEPRA, http://www.coepra.org/) is a modeling competition organized to provide objective testing of various algorithms via the process of blind prediction for chemical, biological, and medical data. COEPRA's stated goals are to advance modeling algorithms and software as well as provide reference datasets to the research community. Transferable Atom Equivalent (TAE) RECON features are electron-density derived descriptors obtained by fragment reconstruction. MOE features are geometrical, structural, physiochemical and topological 2D descriptors.

Low-rank matrices play a central role in large-scale data analysis and dimensionality reduction. They arise in a variety of application areas, among them Principal Component Analysis (PCA), Multidimensional scaling (MDS), Spectral Clustering and related methods, ranking and collaborative filtering, etc. In all these problems, low-rank structure is used to either approximate a general matrix, or to correct for corrupted or missing data. This paper considers the recovery of a low-rank matrix in the simultaneous presence of (a) erasures: most elements are not observed, and (b): errors: among the ones that are observed, a significant fraction at unknown locations are grossly/maliciously corrupted. It is now well recognized that the standard, popular approach to low-rank matrix recovery using SVD as a first step fails spectacularly in this setting [1]. Low-rank matrix completion, which considers only random erasures ([2], [3]) will also fail with even just a few maliciously corrupted entries. In light of this, several recent works have studied an alternate approach based on the natural convex relaxation of minimizing rank plus support. One approach [4], [5] provides deterministic/worst case guarantees for the fully observed setting (i.e.