Sample size and statistical power considerations in high-dimensionality data settings: a comparative study of classification algorithms

no single classifier had optimal performance under all settings. Simulation studies provide useful guidance for the design of biomedical studies involving high-dimensionality data.