Title :
On a Class of Linear Maps for Data Compression
Author :
Kundu, Shovonlal
Author_Institution :
Electronics and Communication Science Unit, Indian Statistical Institute
Abstract :
A method for data compression with linear maps has been developed which is found to produce further reduction in overhead storage requirement, compression/decompression time, and clustering overhead as compared to the affine map method in certain cases. Algorithms have been developed for cluster minimization, cluster identification, and compression matrix calculation that may be applied with advantage in both the methods.
Keywords :
Cluster analysis; computer algebra; data compression; dimensionality reduction; linear transformation; overhead storage; redundancy reduction; Algebra; Clustering algorithms; Costs; Data analysis; Data compression; Minimization methods; Pattern recognition; Vectors; Cluster analysis; computer algebra; data compression; dimensionality reduction; linear transformation; overhead storage; redundancy reduction;
Journal_Title :
Software Engineering, IEEE Transactions on
DOI :
10.1109/TSE.1982.235877
