Interpretable Locally Adaptive Nearest Neighbors

Abstract

When training automated systems, it has been shown to be beneficial to adapt the representation of data by learning a problem-specific metric. This metric is global. We extend this idea and, for the widely used family of k nearest neighbors algorithms, develop a method that allows learning locally adaptive metrics. These local metrics not only improve performance but are naturally interpretable. To demonstrate important aspects of how our approach works, we conduct a number of experiments on synthetic data sets, and we show its usefulness on real-world benchmark data sets.

References in corpus (5)

• Scikit-learn: Machine Learning in Python
• UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction
• European Union regulations on algorithmic decision-making and a "right to explanation"
• A Survey on Metric Learning for Feature Vectors and Structured Data
• Parametric Local Metric Learning for Nearest Neighbor Classification