Fine-tuning Deep Belief Networks using Harmony Search

Autor(es)

Papa, Joao Paulo ^[1] ; Scheirer, Walter ^[2] ; Cox, David Daniel ^[2]

Afiliação do(s) autor(es)

^[1] UNESP Univ Estadual Paulista, Dept Comp, Bauru - Brazil

^[2] Harvard Univ, Ctr Brain Sci, Cambridge, MA 02138 - USA

• Tipo de documento: Artigo Científico
• Fonte: APPLIED SOFT COMPUTING; v. 46, p. 875-885, SEP 2016.
• Citações Web of Science: 32

Resumo

In this paper, we deal with the problem of Deep Belief Networks (DBNs) parameters fine-tuning by means of a fast meta-heuristic approach named Harmony Search (HS). Although such deep learning-based technique has been widely used in the last years, more detailed studies about how to set its parameters may not be observed in the literature. We have shown we can obtain more accurate results comparing HS against with several of its variants, a random search and two variants of the well-known Hyperopt library. The experimental results were carried out in two public datasets considering the task of binary image reconstruction, three DBN learning algorithms and three layers. (C) 2015 Elsevier B.V. All rights reserved. (AU)

• Processo FAPESP: 13/20387-7 - Otimização de hiperparâmetros em arquiteturas de aprendizado em profundidade
• Beneficiário: João Paulo Papa
• Modalidade de apoio: Bolsas no Exterior - Pesquisa
• Processo FAPESP: 14/16250-9 - Sobre a otimização de parâmetros em técnicas de aprendizado de máquina: avanços e paradigmas
• Beneficiário: João Paulo Papa
• Modalidade de apoio: Auxílio à Pesquisa - Regular