Model selection based on particle swarm optimization for omics data classification

A new model selection algorithm based on binary particle swarm optimization is proposed for omics data classification. Particularly, the algorithm is designed to handle the high dimensionality, small sample size and class imbalance problems that are inherent in omics data. The particles encode candidate combinations of data sampling, feature selection, classification models and their corresponding parameter settings. The binary swarm optimization is targeted at the best classification performance. The particle velocity and position are iteratively updated until some stopping iteration is met and the optimal solution model combination is output. The simulative results on eight real-world omics datasets show that the proposed model selection algorithm is capable of avoiding the bias introduced by manual settings and leading to accurate and reliable classification performance.