抄録
We study unique functions of an Autonomous Asynchronous Cooperation (AAC) useful to a Problem-Solving Environment (PSE) for a target class having local minima. In our paper, assume that an asynchronous cooperation system consists of network-linked computers and software packages. Several software packages deploy to network-linked computers to autonomously work toward common goals to solve a problem. The behavior of an AAC is complex and hardly predictable, even if any software does not have randomness. Our study clarified that our mathematical model for an AAC shows stochastic behavior which dynamically changes depending on the status of itself. An AAC is the execution environment itself as the distributed Monte Carlo method. Furthermore, the observed behavior of an AAC applied to solve a typical problem demonstrated a complexity shown in our mathematical model. An AAC effectively provides the unique functions in developing a PSE for a target class having local minima. Then, PSE application developers can concentrate on problem-solving without being bothered by programming technique.
