In the context of ensemble systems, feature selection methods can be used to provide different subsets of attributes for the individual classifiers, aiming to reduce redundancy among the attributes of a pattern and to increase the diversity in such systems. Among the several techniques that have been proposed in the literature, optimization methods have been used to find the optimal subset of attributes for an ensemble system. In this paper, an investigation of two optimization techniques, genetic algorithm and ant colony optimization, will be used to guide the distribution of the features among the classifiers. This analysis will be conducted in the context of heterogeneous ensembles and using different ensemble sizes.

No abstract

No abstract

This paper presents preliminary results of an embedded system development to aid identification and automated counting of scale cochineal (Diaspis echinocacti) in samples of forage palm, within a laboratory environment. Classical techniques of digital image processing are used for segmentation and recognition of cochineal in the third instar phase, with acceptable accuracy compared to visual recognition by a specialist. The complexity of differentiating between males and females of the cochineal and the challenge of classifying the different stages of development, indicates the need to incorporate object recognition techniques and frameworks based on machine learning.Resumo: Este trabalho apresenta resultados preliminares do desenvolvimento de um sistema embarcado para auxílio à identificação e contagem automatizada da cochonilha de escama (Diaspis echinocacti) em amostras de palma forrageira, em ambiente laboratorial. São utilizadas técnicas clássicas de processamento digital de imagens para segmentação e reconhecimento das cochonilhas na fase de terceiro ínstar, com acurácia aceitável frente ao reconhecimento visual por especialista. A complexidade na diferenciação entre machos e fêmeas da cochonilha, bem como o desafio de classificar as diferentes fases de desenvolvimento apontam a necessidade de incorporar técnicas e frameworks de reconhecimento de objetos baseados em aprendizagem de máquina.