Theory Of The GMM Kernel

Li Ping, Zhang Cun-hui. Arxiv 2016

[Paper]    
ARXIV

We develop some theoretical results for a robust similarity measure named “generalized min-max” (GMM). This similarity has direct applications in machine learning as a positive definite kernel and can be efficiently computed via probabilistic hashing. Owing to the discrete nature, the hashed values can also be used for efficient near neighbor search. We prove the theoretical limit of GMM and the consistency result, assuming that the data follow an elliptical distribution, which is a very general family of distributions and includes the multivariate $t$-distribution as a special case. The consistency result holds as long as the data have bounded first moment (an assumption which essentially holds for datasets commonly encountered in practice). Furthermore, we establish the asymptotic normality of GMM. Compared to the “cosine” similarity which is routinely adopted in current practice in statistics and machine learning, the consistency of GMM requires much weaker conditions. Interestingly, when the data follow the $t$-distribution with $\nu$ degrees of freedom, GMM typically provides a better measure of similarity than “cosine” roughly when $\nu <8$ (which is already very close to normal). These theoretical results will help explain the recent success of GMM in learning tasks.

Similar Work

• Generalized Intersection Kernel
• Multivariate Representation Learning For Information Retrieval
• Expanding The Family Of Grassmannian Kernels An Embedding Perspective
• Locality-sensitive Hashing For F-divergences Mutual Information Loss And Beyond