Rafael Cabañas

AMIDST: A Java toolbox for scalable probabilistic machine learning

Abstract The AMIDST Toolbox is an open source Java software for scalable probabilistic machine learning with a special focus on (massive) streaming data. The toolbox supports a flexible modelling language based on probabilistic graphical models with latent variables. AMIDST provides parallel and distributed implementations of scalable algorithms for doing probabilistic inference and Bayesian parameter learning in the specified models. These algorithms are based on a flexible variational message passing scheme, which supports discrete and continuous variables from a wide range of probability distributions.

Financial Data Analysis with PGMs Using AMIDST

The AMIDST Toolbox an open source Java 8 library for scalable learning of probabilistic graphical models (PGMs) based on both batch and streaming data. An important application domain with streaming data characteristics is the banking sector, where we may want to monitor individual customers (based on their financial situation and behavior) as well as the general economic climate. Using a real financial data set from a Spanish bank, we have previously proposed and demonstrated a novel PGM framework for performing this type of data analysis with particular focus on concept drift. The framework is implemented in the AMIDST Toolbox, which was also used to conduct the reported analyses. In this paper, we provide an overview of the toolbox and illustrate with code examples how the toolbox can be used for setting up and performing analyses of this particular type.

On SPI-Lazy Evaluation of Influence Diagrams

Influence Diagrams are an effective modelling framework for analysis of Bayesian decision making under uncertainty. Improving the performance of the evaluation is an element of crucial importance as real-world decision problems are more and more complex. Lazy Evaluation is an algorithm used to evaluate Influence Diagrams based on message passing in a strong junction tree. This paper proposes the use of Symbolic Probabilistic Inference as an alternative to Variable Elimination for computing the clique-to-clique messages in Lazy Evaluation of Influence Diagrams.

On SPI for Evaluating Influence Diagrams

An Influence Diagram is a probabilistic graphical model used to represent and solve decision problems under uncertainty. Its evaluation requires to perform a series of combinations and marginalizations with the potentials attached to the Influence Diagram. Finding an optimal order for these operations, which is NP-hard, is an element of crucial importance for the efficiency of the evaluation. The SPI algorithm considers the evaluation as a combinatorial factorization problem. In this paper, we describe how the principles of SPI can be used to solve Influence Diagrams. We also include an evaluation of different combination selection heuristics and a comparison with the variable elimination algorithm.

Using Binary Trees for the Evaluation of Influence Diagrams

This paper proposes the use of binary trees for representing and managing the potentials involved in Influence Diagrams. This kind of tree allows representing context-specific independencies that are finer-grained compared to those encoded using other representations. This enhanced capability can be used to improve the efficiency of the inference algorithms used for Influence Diagrams. Moreover, binary trees allow computing approximate solutions when exact inference is not feasible. In this work we describe how binary trees can be used to perform this approximate evaluation and we compare them with other structures present in the literature.

Variable Elimination for Interval-Valued Influence Diagrams

Influence diagrams are probabilistic graphical models used to represent and solve decision problems under uncertainty. Sharp numerical values are required to quantify probabilities and utilities. Yet, real models are based on data streams provided by partially reliable sensors or experts. We propose an interval-valued quantification of these parameters to gain realism in the modelling and to analyse the sensitivity of the inferences with respect to perturbations of the sharp values. An extension of the classical influence diagrams formalism to support interval-valued potentials is provided. Moreover, a variable elimination algorithm especially designed for these models is developed and evaluated in terms of complexity and empirical performances.

Approximate Lazy Evaluation of Influence Diagrams

Influence Diagrams are a tool used to represent and solve decision problems under uncertainty. One of the most efficient exact methods used to evaluate Influence Diagrams is Lazy Evaluation. This paper proposes the use of trees for representing potentials involved in an Influence Diagram in order to obtain an approximate Lazy Evaluation of decision problems. This method will allow to evaluate complex decision problems that are not evaluable with exact methods due to their computational cost. The experimental work compares the efficiency and goodness of the approximate solutions obtained using different kind of trees.

Virtual Subconcept Drift Detection in Discrete Data Using Probabilistic Graphical Models

A common problem in mining data streams is that the distribution of the data might change over time. This situation, which is known as concept drift, should be detected for ensuring the accuracy of the models. In this paper we propose a method for subconcept drift detection in discrete streaming data using probabilistic graphical models. In particular, our approach is based on the use of conditional linear Gaussian Bayesian networks with latent variables. We demonstrate and analyse the proposed model using synthetic and real data.

Heuristics for determining the elimination ordering in the influence diagram evaluation with binary trees

Finding an optimal elimination ordering is a NP-hard problem of crucial importance for the efficiency of the Influence Diagrams evaluation. Some of the traditional methods for determining the elimination ordering use heuristics that consider that potentials are represented as tables. However, if potentials are represented using binary trees traditional methods may not offer the best results. In the present paper, two new heuristics that consider that potentials are represented as binary trees are proposed. As a result, the storage requirements for evaluating an ID with binary trees is reduced.

Evaluating asymmetric decision problems with binary constraint trees

This paper proposes the use of binary trees in order to represent and evaluate asymmetric decision problems with Influence Diagrams (IDs). Constraint rules are used to represent the asymmetries between the variables of the ID. These rules and the potentials involved in IDs will be represented using binary trees. The application of these rules can reduce the size of the potentials of the ID. As a consequence the efficiency of the inference algorithms will be improved.