The Fuzzy Learning Vector Quantization with Allied Fuzzy C-means Clustering for Clustering Noisy Data

摘要

Fuzzy Learning Vector Quantization (FLVQ) is the integration of Learning Vector Quantization (LVQ) and Fuzzy C-means (FCM) clustering, i.e. FLVQ is a FCM-based model. FCM is sensitive to noises, so FLVQ has the same noise sensitivity problem as FCM. To solve the noise sensitivity problem of FLVQ, a novel fuzzy learning vector quantization model, called Allied Fuzzy Learning Vector Quantization (AFLVQ), is proposed. AFLVQ integrates learning vector quantization and Allied Fuzzy C-means clustering (AFCM) and uses the membership and typicality values from AFCM as learning rates. AFLVQ can produce memberships and possibilities simultaneously, and it overcomes the noise sensitivity shortcoming of FLVQ. Experiments show the better performances of AFLVQ.