How far you can get using machine learning black-boxes
Nenhuma Miniatura disponível
Data
2010-12-01
Autores
Rocha, Anderson
Papa, João Paulo [UNESP]
Meira, Luis A. A.
Título da Revista
ISSN da Revista
Título de Volume
Editor
Resumo
Supervised Learning (SL) is a machine learning research area which aims at developing techniques able to take advantage from labeled training samples to make decisions over unseen examples. Recently, a lot of tools have been presented in order to perform machine learning in a more straightforward and transparent manner. However, one problem that is increasingly present in most of the SL problems being solved is that, sometimes, researchers do not completely understand what supervised learning is and, more often than not, publish results using machine learning black-boxes. In this paper, we shed light over the use of machine learning black-boxes and show researchers how far they can get using these out-of-the-box solutions instead of going deeper into the machinery of the classifiers. Here, we focus on one aspect of classifiers namely the way they compare examples in the feature space and show how a simple knowledge about the classifier's machinery can lift the results way beyond out-of-the-box machine learning solutions. © 2010 IEEE.
Descrição
Palavras-chave
K-Nearest neighbors, Machine learning black-boxes, Metrics space, Neural networks, Optimum-Path forest, Pattern analysis, Support vector machines
Como citar
Proceedings - 23rd SIBGRAPI Conference on Graphics, Patterns and Images, SIBGRAPI 2010, p. 193-200.