Modularization in Belief-Desire-Intention agent programming and artifact-based environments.

This article proposes an extension for the Agents and Artifacts meta-model to enable modularization. We adopt the Belief-Desire-Intention (BDI) model of agency to represent independent and reusable units of code by means of modules. The key idea behind our proposal is to take advantage of the syntactic notion of namespace, i.e., a unique symbol identifier to organize a set of programming elements. On this basis, agents can decide in BDI terms which beliefs, goals, events, percepts and actions will be independently handled by a particular module. The practical feasibility of this approach is demonstrated by developing an auction scenario, where source code enhances scores of coupling, cohesion and complexity metrics, when compared against a non-modular version of the scenario. Our solution allows to address the name-collision issue, provides a use interface for modules that follows the information hiding principle, and promotes software engineering principles related to modularization such as reusability, extensibility and maintainability. Differently from others, our solution allows to encapsulate environment components into modules as it remains independent from a particular BDI agent-oriented programming language.

Wearables for Engagement Detection in Learning Environments: A Review.

Appropriate teaching-learning strategies lead to student engagement during learning activities. Scientific progress and modern technology have made it possible to measure engagement in educational settings by reading and analyzing student physiological signals through sensors attached to wearables. This work is a review of current student engagement detection initiatives in the educational domain. The review highlights existing commercial and non-commercial wearables for student engagement monitoring and identifies key physiological signals involved in engagement detection. Our findings reveal that common physiological signals used to measure student engagement include heart rate, skin temperature, respiratory rate, oxygen saturation, blood pressure, and electrocardiogram (ECG) data. Similarly, stress and surprise are key features of student engagement.

How good is crude MDL for solving the bias-variance dilemma? An empirical investigation based on Bayesian networks.

The bias-variance dilemma is a well-known and important problem in Machine Learning. It basically relates the generalization capability (goodness of fit) of a learning method to its corresponding complexity. When we have enough data at hand, it is possible to use these data in such a way so as to minimize overfitting (the risk of selecting a complex model that generalizes poorly). Unfortunately, there are many situations where we simply do not have this required amount of data. Thus, we need to find methods capable of efficiently exploiting the available data while avoiding overfitting. Different metrics have been proposed to achieve this goal: the Minimum Description Length principle (MDL), Akaike's Information Criterion (AIC) and Bayesian Information Criterion (BIC), among others. In this paper, we focus on crude MDL and empirically evaluate its performance in selecting models with a good balance between goodness of fit and complexity: the so-called bias-variance dilemma, decomposition or tradeoff. Although the graphical interaction between these dimensions (bias and variance) is ubiquitous in the Machine Learning literature, few works present experimental evidence to recover such interaction. In our experiments, we argue that the resulting graphs allow us to gain insights that are difficult to unveil otherwise: that crude MDL naturally selects balanced models in terms of bias-variance, which not necessarily need be the gold-standard ones. We carry out these experiments using a specific model: a Bayesian network. In spite of these motivating results, we also should not overlook three other components that may significantly affect the final model selection: the search procedure, the noise rate and the sample size.