Matrix-Based Evolutionary Computation

Computational intelligence (CI), including artificial neural network, fuzzy logic, and evolutionary computation (EC), has rapidly developed nowadays. Especially, EC is a kind of algorithm for knowledge creation problem solving, playing significant role in CI (AI). However, traditional algorithms have faced great challenge heavy computational burden long running time large-scale (e.g., with many variables) problems. How to efficiently extend solve complex problems become one the most research topics AI communities. To this aim, paper proposes matrix-based (MEC) framework solving or super optimization The proposed an entirely new perspective on algorithm, from solution representation operators. In framework, whole population (containing set individuals) defined as matrix, where row stands individual column dimension (decision variable). This way, parallel computing functionalities matrix can be directly easily carried out high performance resources accelerate speed gives two typical examples MEC algorithms, named genetic particle swarm optimization. Their representations are presented operators based described. Moreover, complexity analyzed experiments conducted show that these efficient reducing large scale decision variables. promising way big data environment, leading direction AI.