Multiple machine learning algorithms, validation of external clinical cohort and assessments of model gain effects will better serve cancer research on bioinformatic models

Abstract

Abstract Bioinformatics models greatly contribute to individualized assessments of cancer patients. However, considerable research neglected some critical technological points, including comparisons of multiple modeling algorithms, evaluating gain effects of constructed model, comprehensive bioinformatics analyses and validation of clinical cohort. These issues are worthy of emphasizing, which will better serve future cancer research.

Keywords

• Machine learning algorithm
• Gain effect
• Bioinformatics
• Prognostic model
• Clinical validation