Immune Plasma Programming: A new evolutionary computation-based automatic programming method

Abstract

Immune plasma therapy, one of the treatment modalities, has proven effective in combating the now rapidly spreading COVID-19 and many other pandemics. The immune plasma algorithm (IPA), inspired by the application phases of this treatment modality, is a recently proposed metaheuristic algorithm. Since its introduction, it has achieved promising results in engineering applications. In this paper, we propose for the first time immune plasma programming (IPP) based on the structure of IPA as a new evolutionary computation-based automatic programming (AP) method. It is compared with well-known AP methods such as artificial bee colony programming, genetic programming, and cartesian ant programming using symbolic regression test problems. It is also compared with baseline methods, many of which are based on recurrent neural networks and a real-word problem is solved. The control parameters of IPP are also tuned separately. The results of the experiments and statistical tests have shown that the prediction accuracy and convergence speed of the models produced by IPP are high. Therefore, IPP has been proposed as a method that can be used to solve various problems.