Automatic Adjustment of Discriminant Adaptive Nearest Neighbor

K-Nearest Neighbors relies on the definition of a global metric. In contrast, Discriminant Adaptive Nearest Neighbor (DANN) computes a different metric at each query point based on a local Linear Discriminant Analysis. In this paper, we propose a technique to automatically adjust the hyper-parameters in DANN by the optimization of two quality criteria. The first one measures the quality of discrimination, while the second one maximizes the local class homogeneity. We use a Bayesian formulation to prevent overfitting.

• Type: Conference
• Year: 2006
• Where: ICPR
• Authors: Cédric Archambeau, Michel Verleysen, Nicolas Delannay
• Keywords: Computer Vision, Discriminant Adaptive Nearest, ICPR 2006, Linear Discriminant Analysis, Local Class Homogeneity