Adoption of Combinatorial Graph for Inhibitory Process in Optimization Problems

摘要

This work is the follow up of the 'Process of Inhibition' supporting neural network models constructed based on biological theories in the earlier works to perform successfully in solving the optimization type problems. The algorithmic implementation of process of inhibition is replaced using the proposed Combinatorial Inhibition Graph (CIG) in this work. This adopted graph produces better performance than the algorithmic implementation done in the earlier works and is found to be suitable for further neural implementation for the process of inhibition. The combinatorial graph which is also suggested suitable for achieving improved performance in a concurrent processing environment is appropriately adopted in this work. The implemented CIG was not only combinatorics in nature but also act as a multi relational graph in solving complex problem like sum-of-subset.