Vicenç Soler

Rectangular basis functions applied to imbalanced datasets

Rectangular Basis Functions Networks (RecBFN) come from RBF Networks, and are composed by a set of Fuzzy Points which describe the network. In this paper, a set of characteristics of the RecBF are proposed to be used in imbalanced datasets, especially the order of the training patterns. We will demonstrate that it is an important factor to improve the generalization of the solution, which is the main problem in imbalanced datasets. Finally, this solution is compared with other important methods to work with imbalanced datasets, showing our method works well with this type of datasets and that an understandable set of rules can be extracted.

Imbalanced Datasets Classification by Fuzzy Rule Extraction and Genetic Algorithms

We propose a method based on the extraction of fuzzy rules by genetic algorithms for the classification of imbalanced datasets when understandability is an issue. We propose a new method for fuzzy variable construction based on modifying the set of fuzzy variables obtained by the RecBF/DDA algorithm. Later, these variables are recombined to obtain fuzzy rules by means of a genetic algorithm. The method has been developed for the detection of Downs syndrome in fetus. We provide empirical results showing its accuracy for this task. Furthermore, we provide more generic experimental results over UCI datasets proving that the method can have a wider applicability

Fuzzy rule extraction using recombined RecBF for very-imbalanced datasets

An introduction to how to use RecBF to work with very-imbalanced datasets is described. In this paper, given a very-imbalanced dataset obtained from medicine, a set of Membership Functions (MF) and Fuzzy Rules are extracted. The core of this method is a recombination of the Membership Functions given by the RecBF algorithm which provides a better generalization than the original one. The results thus obtained can be interpreted as sets of low number of rules and MF.

Extracting a Fuzzy System by Using Genetic Algorithms for Imbalanced Datasets Classification: Application on Down’s Syndrome Detection

This chapter presents a new methodology to extract a Fuzzy System by using Genetic Algorithms for the classification of imbalanced datasets when the intelligibility of the Fuzzy Rules is an issue. We propose a method for fuzzy variable construction, based on modifying the set of fuzzy variables obtained by the DDA/RecBF clustering algorithm. Afterwards, these variables are recombined to obtain Fuzzy Rules by means of a Genetic Algorithm. The method has been developed for the prenatal Down’s syndrome detection during the secondtrimester of pregnancy. We present empirical results showing its accuracy for this task. Furthermore, we provide more generic experimental results over UCI datasets proving that the method can have a wider applicability on imbalanced datasets.

Adapting Fuzzy Points for Very-Imbalanced Datasets

RecBF networks come from RBF Networks, and are composed by a set of fuzzy points which describe the network. How to adapt these fuzzy points to work with very imbalanced datasets is described in this paper. Once the fuzzy points are found and adapted by the modifications proposed, they are included in a fuzzy system that classifies imbalanced datasets. The results showed will proof empirically that the proposed changes work well for very-imbalanced datasets

Use of EAC in learning digital systems

In electronic circuit subjects it is very important that students practice their exercise with analysis and synthesis of circuits in order to acquire knowledge, skill and competences. Usually, students are poorly motivated and thus problemspsila classes get reduced to the professor giving the solution to the set problems. Studentspsila only mission is then to copy the information from the blackboard without having to study, develop or think about the problem. In this paper, we present the use of an interactive docent tool, Educlick, which has been adapted to these problemspsila classes . Educlick is based on the use of electronic answer remote controls (clickers). The classes are wholly directed. The experience is done in the subject of Digital Systems of Computer Science at the Universitat Autnoma of Barcelona. The objective is to increment studentspsila participation in class and thus improve their learning of the design of combinational and sequential digital systems.

Digital circuit simulator via Internet

In this article, the authors present an application aimed at students who attend the subject of Digital Systems and want to make a continuous and progressive study of it. Thats why a logical simulator via Internet is being developed within the Internet portal of the subject of Digital Systems. It allows students to implement and to verify the practical part of the subject before the laboratory session. This simulator consists of an oscilloscope, a function generator, a power supply, in addition to several integrated circuits, which can be either combinational or sequential, or both.