Weighted Ensemble for Plant Protein Subcellular Localization Using Particle Swarm Optimization

摘要

This article presents the application of particle swarm optimization (PSO) for constructing a weighted vote ensemble to identify localizations of proteins in plant cells. The procedures were conducted using robotic process automation (RPA) to collect results from many existed classifiers as inputs of the ensemble and the results were combined using the optimized weighted to conclude the final decision. The constructed web-based application can help facilitates the biological researchers and scientists to recognize not only the primary location of the proteins in plant cells but also the alternative location in which such proteins can possibly be founded. From the experiment results, the PSO-weighted vote ensemble outperforms the majority vote ensemble with up to 92% accuracy.