Simulating the Cognitive Thought Process with a Modified Genetic Algorithm

Abstract

Cognitive Computing is a developing paradigm with its applications found in almost every field. It aims to develop computing methodologies and systems inspired by mind’s capabilities. A thought is the most fundamental capability of the mind. Hence it is important in the cognitive computing to understand the Thought Process. The purpose of this paper is to develop a computer model that simulates the Thought Process. The paper proposes that the Genetic Algorithm (GA) can be used for the same. Both cognitive model and computer model of the thought process with the help of GA has been given. In the computer model, a modification of GA has been implemented, which consists of a new crossover operator called Learning Crossover operator. The new crossover operator is not a replacement but is a supplement to the existing crossover operators. The GA is implemented over the Travelling Salesperson Problem (TSP), which is a classical NP problem and hence the algorithm is possible to be implemented on any other problem. The modification to the GA aims to improve the Thought Process Simulation. But it can also improve the performance of GA, when further explored.