FoMo vs. Final Classifier Performance Comparison

This dataset compares the performance of the FoMo feature selection method against the final classifier developed on the 55-feature set versus the 38-feature set from the original study. It reports the initial and final (post feature pruning and oversampling) modeling results and compares them to FoMo v1.0 and v1.3. A difference column compares performance between v1.3 and the final classifier from the 55-feature set.

Use Cases

• Benchmarking the FoMo feature selection method against other techniques.
• Studying the impact of feature pruning and oversampling on classifier performance.
• Comparing model versions (v1.0, v1.3) within the FoMo framework.

Strengths

• Direct comparison between two feature set sizes (55 vs. 38).
• Includes both initial and final modeling results after preprocessing steps.
• Contains a specific performance difference metric between key model versions.

Limitations

• Dataset size is very small (5.5 KB), limiting statistical power.
• Row and column details are not provided in the description.
• Freshness is unknown; the underlying study dates to 2026.

Provenance

Source

Figshare repository for the original research.

Collection Method

Performance metrics computed from experimental comparison.

Time Range

Not specified.

Geography

Not specified.