Cost-Sensitive Hypergraph Learning With F-Measure Optimization.

Area of Science:

• Machine Learning
• Data Science
• Artificial Intelligence

Background:

• Data imbalance is a prevalent challenge in machine learning, where certain classes have significantly fewer samples.
• Existing cost-sensitive learning methods often struggle with precise cost value determination, even with domain expertise.

Purpose of the Study:

• To propose a novel cost-sensitive hypergraph learning method for effectively addressing data imbalance.
• To leverage the F-measure for optimizing cost information in imbalanced data classification.

Main Methods:

• Utilizing hypergraph structures to capture high-order relationships within imbalanced datasets.
• Employing F-measure optimization to derive accurate cost information.
• Implementing cost-sensitive hypergraph learning with the optimized cost values.

Main Results:

• The proposed method demonstrates effectiveness in handling imbalanced data classification tasks.
• Experimental results validate the superiority of the F-measure optimized cost-sensitive hypergraph learning approach.

Conclusions:

• The developed cost-sensitive hypergraph learning method offers a robust solution for imbalanced data problems.
• Optimizing cost information using F-measure enhances the performance of hypergraph learning on imbalanced datasets.