Félix Castro

Applying Data Mining Techniques to e-Learning Problems

This chapter aims to provide an up-to-date snapshot of the current state of research and applications of Data Mining methods in e-learning. The cross-fertilization of both areas is still in its infancy, and even academic references are scarce on the ground, although some leading education-related publications are already beginning to pay attention to this new field. In order to offer a reasonable organization of the available bibliographic information according to different criteria, firstly, and from the Data Mining practitioner point of view, references are organized according to the type of modeling techniques used, which include: Neural Networks, Genetic Algorithms, Clustering and Visualization Methods, Fuzzy Logic, Intelligent agents, and Inductive Reasoning, amongst others. From the same point of view, the information is organized according to the type of Data Mining problem dealt with: clustering, classification, prediction, etc.

Advances in Soft Computing and Its Applications

Genetic Algorithms (GAs) have long been recognized as powerful tools for optimization of complex problems where traditional techniques do not apply. However, although the convergence of elitist GAs to a global optimum has been mathematically proven, the number of iterations remains a case-by-case parameter. We address the problem of determining the best GA out of a family of structurally different evolutionary algorithms by solving a large set of unconstrained functions. We selected 4 structurally different genetic algorithms and a non-evolutionary one (NEA). A schemata analysis was conducted further supporting our claims. As the problems become more demanding, the GAs significantly and consistently outperform the NEA. A particular breed of GA (the Eclectic GA) is superior to all other, in all cases.

Rule-Based Assistance to Brain Tumour Diagnosis Using LR-FIR

This paper describes a process of rule-extraction from a multi-centre brain tumour database consisting of nuclear magnetic resonance spectroscopic signals. The expert diagnosis of human brain tumours can benefit from computer-aided assistance, which has to be readily interpretable by clinicians. Interpretation can be achieved through rule extraction, which is here performed using the LR-FIR algorithm, a method based on fuzzy logic. The experimental results of the classification of three groups of tumours indicate in this study that just three spectral frequencies, out of the 195 from a range pre-selected by experts, are enough to represent, in a simple and intuitive manner, most of the knowledge required to discriminate these groups.

Genetic Learning of Fuzzy Parameters in Predictive and Decision Support Modelling

Abstract In this research a genetic fuzzy system (GFS) is proposed that performs discretization parameter learning in the context of the Fuzzy Inductive Reasoning (FIR) methodology and the Linguistic Rule FIR (LR-FIR) algorithm. The main goal of the GFS is to take advantage of the potentialities of GAs to learn the fuzzification parameters of the FIR and LR-FIR approaches in order to obtain reliable and useful predictive (FIR) models and decision support (LR-FIR) models. The GFS is evaluated in an e-learning context.

Causal relevance to improve the prediction accuracy of dynamical systems using inductive reasoning

In this paper, the concept of causal relevance (CR) is introduced in the context of the fuzzy inductive reasoning (FIR) modelling and simulation methodology. The idea behind CR is to quantify how much influence each system variable has, from the spatial and temporal points of view, on the prediction of the output. This paper introduces the FIR inference engine, and describes how it can be improved by means of the CR concept, helping to reduce uncertainty during the forecasting stage. The FIR inference engine is based on the k-nearest neighbour classification rule, commonly used in the field of pattern recognition, and uses a Euclidean distance measure to compute the distance between neighbours. In this paper, a weight-Euclidean distance measure is proposed that is able to find better quality neighbours by using the CR concept. Applications from different fields are studied in the light of the prediction process, and a comparison between the accuracy of the predictions obtained when using the classical inference engine and the CR option is performed. The results obtained from this research show that FIR predictions are more accurate and precise when the CR option is used, especially for systems where classical FIR forecasting performs rather poorly.

Data Mining of Virtual Campus Data

As mentioned elsewhere in this book, e-learning offers “a new context for education where large amounts of information describing the continuum of the teaching–learning interactions are endlessly generated and ubiquitously available”. But raw information by itself may be of no help to any of the e-learning actors. The use of Data Mining methods to extract knowledge from this information can, therefore, be an adequate approach to follow in order to use the obtained knowledge to fit the educational proposal to the students’ needs and requirements. This chapter provides a case study in which several advanced Data Mining techniques are employed to extract different types of knowledge from virtual campus data concerning students system usage behaviour. The diverse palette of Data Mining problems addressed here include data clustering and visualization, outlier detection, classification, feature selection, and rule extraction. They concern diverse e-learning problems, such as the characterization of atypical students’ behaviour and the prediction of students’ performance.

Causal Relevancy Approaches to Improve the Students' Prediction Performance in an e-Learning Environment

In this work, four different causal relevancy (CR) approaches are implemented within the inference engine of the fuzzy inductive reasoning (FIR) methodology. The idea behind CR is to quantify how much influence each system feature has, on the forecasting of the output. This paper presents and discusses the FIR inference engine, and describes how it can be enhanced using the causal relevancy methods proposed in this study. The first two CR methods compute the relevancy of each feature by means of the quality of the optimal mask, obtained in the qualitative model identification step of the FIR methodology. The last two CR methods are based on the prediction error of a validation data set, not used in the model identification process. The CR approaches presented in the paper are applied to a real e-learning course with the goal of improve studentspsila behavior predictions. The experiments carried out with the available data indicate that lower prediction errors are obtained using the CR approaches when compared with the results obtained by the classical FIR inference engine. The new approaches help to improve the understanding of the educative process by describing how much influence each system feature has on the output.

Advances in Artificial Intelligence and Its Applications

We present an extension of GLukG, a logic that was introduced in [8] as a three-valued logic under the name of G3. GLukG is a paraconsistent logic defined in terms of 15 axioms, which serves as the formalism to define the p-stable semantics of logic programming. We introduce a new axiomatic system, N-GLukG, a paraconsistent logic that possesses strong negation. We use the 5-valued logic N ′ 5, which is a conservative extension of GLukG, to help us to prove that N-GLukG is an extension of GLukG. N-GLukG can be used as the formalism to define the p-stable semantics as well as the stable semantics.