Search

In general, computer simulations are a useful means to confirm the effectiveness of a control scheme, but the actual system is not so easy to handle. In order to overcome such underlying realistic problems, adaptability of neural networks is focused on by many researchers. In this study, an on-line training model for self-turning control is constructed based on the famous feedback error learning model, and its effectiveness is investigated through a task for controlling an inverted pendulum. As a result of several experiments, repeatable strange phenomena, such as seasonal variation of training period and reversalization of control command, are observed.

A Hopfield network is a good tool for solving combinatorial optimization problems. Since its proposal, research activity around this area has been accelerated with the help of related ideas. But some of them are just techniques to get better solutions, so no meaningful phenomena might be observed in the field of the actual spin glass. In this paper, therefore, another search method inspired from an analogy between the Hopfield network and the spin glass is proposed. It is quite a simple idea to control a threshold of the Hopfield network corresponding to a magnetic field in the spin glass. In order to confirm its effectiveness, an N queens problem is adopted. As a result, it is found experimentally that the proposed method shows better score, e.g. a ratio of correct answer and an averaged final energy, than the conventional one.

In general, a biological nervous system consists of two major parts, i.e., a motor control subsystem and a sensory reception subsystem, and they work together cooperatively as direct-inverse transformation. Paying attention to this bi-directional computing manner, it is essential to perform various kinds of complex activities. Also, it must be important in most application-oriented tasks. In order to confirm its effectiveness, a bi-directional computing architecture inspired from above-mentioned scheme is constructed and applied to time series prediction tasks. At that time, one subsystem in the bi-directional model is assigned to ordinary future prediction (present→future), while the other is to past prediction (present→past) newly introduced in this study. As a result of computer simulations, it is clear that the proposed bi-directional computing architecture shows better performance than the conventional uni-directional one.