Small groups in multidimensional feature space: Two examples of supervised two-group classification from biomedicine

Area of Science:

• Biomedical data analysis
• Statistical modeling
• Machine learning

Background:

• Biomedical datasets often have limited samples and numerous features, posing analytical challenges like overfitting.
• Accurate analysis is crucial for reliable interpretation and clinical decision-making.

Purpose of the Study:

• To provide a tutorial on a mathematical approach for supervised two-group classification with small sample, high-feature datasets.
• To demonstrate methods for handling data uncertainties and missing values without data imputation or removal.

Main Methods:

• Utilized a supervised two-group classification framework on medical datasets.
• Developed a method incorporating feature distribution size/shape and pairwise feature relations as derived prognostic factors.
• Implemented similarity calculations accounting for within-group feature variation and measurement inaccuracies.

Main Results:

• The described mathematical approach enhances accuracy and reliability in analyzing small, high-dimensional biomedical data.
• The tutorial effectively addresses uncertainties and missing values, preserving data integrity.

Conclusions:

• This method offers a robust solution for challenging biomedical data analysis scenarios.
• Improved analytical techniques are vital for advancing precision medicine and diagnostics.