In this paper, we propose a new reinforcement learning scheme called CHQ that could efficiently acquire appropriate policies under partially observable Markov decision processes (POMDP) involving probabilistic state transitions, that frequently occurs in multi-agent systems in which each agent independently takes a probabilistic action based on a partial observation of the underlying environment. A key idea of CHQ is to extend the HQ-learning proposed by Wiering et al. in such a way that it could learn the activation order of the MDP subtasks as well as an appropriate policy under each MDP subtask. The goodness of the proposed scheme is experimentally evaluated. The result of experiments implies that it can acquire a deterministic policy with a sufficiently high success rate, even if the given task is POMDP with probabilistic state transitions.

The preparation of magnetic semiconductor thick film (MST) by means of spray printing and application to a temperature/gas/essence sensor have been proposed. The MST pattern is composed of ferrite, ruthenium compound, carbon black, binder and solvent. After the mixed mgnetic semiconductor fluid is sprayed on a substrate, the sample is sintered at 750. The MST with thickness of 40 µm is printed on the substrate in various shapes such as a plate, a ring or a rod. The magnetic property of MST depends on temperature, and the electrical property responds to gas and natural/artificial fruit essence. Therefore, the multipore ceramic MST operates as a gas sensor with high sensitivity and high stability.