Neural networks for feature-extraction in multi-target classification

• Ano de defesa: 2020
• Autor(a) principal: Cambuí, Brendon Gouveia
• Orientador(a): Cerri, Ricardo
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação - PPGCC
• Palavras-chave em Inglês: Multi-Target Classification; Auto-encoders; Restricted Boltzmann Machine; Feature-extraction; Dimensionality reduction
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufscar.br/handle/ufscar/13795

Multi-target learning is a prediction task where each data example is associated with multiple target-variables (outputs) simultaneously. One of the challenges in this research field is related to the high dimensionality of data present in multi-target datasets, and also the high number of target variables having dependencies among themselves. In such scenarios, it is crucial to extract lower-dimensional representations from the original input-space, such that these can be provided as input to other multi-target predictors. In this research, we proposed the use of Auto-Encoders and Restricted Boltzmann Machines as feature extractors in several multi-target classification datasets publicly available. Results were evaluated considering state-of-the-art multi-target classification methods and evaluation measures in the literature. The experiments showed that the neural networks were able to keep the predictive performance even when the extracted features corresponded to a dimension size equivalent to 10% of the original number of features and, in some cases, getting better results than the original datasets.