A three-stage gene selection algorithm based on intrinsic dimension and the concise particle swarm optimization

Key Points

• The algorithm enhances gene selection through a three-stage process, improving the identification of relevant biomarkers.
• Key evidence shows that this approach can effectively reduce data dimensionality using intrinsic dimension techniques.
• Analysis employs particle swarm optimization to refine the selection of genetic features across various datasets and conditions.
• The findings support improved accuracy in biomarker discovery, although further validation is essential on larger datasets.