Aprendizagem por Reforço Hierárquica e Computação Paralela Aplicada ao Problema dos K-servos (Hierarchical Reinforcement Learning and Parallel Computing Applied to the k-server Problem)

Mademerson Leandro Costa (mademersonleandro@uern.br)1, Carlos Alberto Araújo Padilha (caapadilha@inf.ufrgs.br)2, Jorge Dantas Melo (jdmelo@dca.ufrn.br)3, Adrião Duarte Dória Neto (adriao@dca.ufrn.br)3

1Universidade do Estado do Rio Grande do Norte
2Universidade Federal do Rio Grande do Sul
3Universidade Federal do Rio Grande do Norte

This paper appears in: Revista IEEE América Latina

Publication Date: Oct. 2016
Volume: 14,   Issue: 10 
ISSN: 1548-0992

In this paper was proposed an algorithm based on Hierarchical Reinforcement Learning (HRL) and Parallel Computing to solve an online computing problem, the K-Server Problem (KSP). The size of the storage structure used for reinforcement learning to obtain the optimal policy grows exponentially with the number of states and actions, limiting its use to smaller problems due to the curse of dimensionality. The problem is modeled as a multiple steps decision process computed in parallel by applying the Q-learning algorithm to obtain optimal policies in a reduced number of nodes obtained from an clustering process. The results show the applicability of the proposed method to real problems of large size.

Index Terms:
Metrical Task Systems,The K-Server Problem, Curse of Dimensionality, Hierarchical Reinforcement Learning, Q-Learning Algorithm, Parallel Computing.   

Documents that cite this document
This function is not implemented yet.

[PDF Full-Text (274)]