Machine Learning
Size of the dictionary in matching pursuit algorithm
IEEE Transactions on Signal Processing
Kernel matching reduction algorithms for classification
RSKT'08 Proceedings of the 3rd international conference on Rough sets and knowledge technology
Hi-index | 0.00 |
This paper presents a method for dictionary training of Kernel matching pursuits (KMP) [1] applied in large size data classification. This algorithm uses the existing fuzzy clustering technique to design function dictionary from a set of training data. The motivation is to enhance the local and quickly searching ability of basic matching pursuit algorithm. As a result, this method reduces computation complexity of matching pursuit algorithm. Simulation results show the proposed algorithm is efficacious for training large size pattern space.