A Neighborhood Model with Both Distance and Quantity Constraints for Multilabel Data

Area of Science:

• Machine Learning
• Data Mining
• Pattern Recognition

Background:

• Multilabel classification is crucial for complex data analysis.
• Existing algorithms face challenges in balancing accuracy and computational efficiency.

Purpose of the Study:

• To propose a novel distance-based multilabel classification algorithm.
• To improve classification accuracy and computational speed through double constraints.

Main Methods:

• The algorithm combines k-nearest neighbors (kNN) with a neighborhood classifier (NC).
• It introduces a radius constraint in kNN for accuracy and a quantity constraint (k) in NC for speed.
• Bayesian rule is used to estimate label probabilities from constrained neighbors.

Main Results:

• Experimental results demonstrate slight advantages over similar algorithms.
• The proposed method shows improvements in both calculation speed and classification accuracy.

Conclusions:

• The novel algorithm effectively balances computational efficiency and classification performance.
• This approach offers a promising solution for distance-based multilabel classification tasks.