Research on Distributed Stochastic Large-Scale Optimization Method with Learning

点击翻译按钮获取中文摘要

We studied distributed methods for large-scale combinatorial optimization problems.(1) Based on biological evolution which means the adaptation (learning) to the enviromment, we modified simulated annealing and developed a new stochastic optimization method.(2) We studied optimization without cost function using learning of neural networks.The application of the developed methods to layout design of VLSI, i. e., placement and routing problems, showed the effectiveness.