COMPRESSION OF 3-MODE DATA ARRAYS BY B-SPLINES PRIOR TO 3-MODE PRINCIPAL COMPONENT ANALYSIS

Three-mode PCA is very computer demanding. It requires a large amount of storage space and many floating point operations (FLOPS). By using three-mode B-spline compression of three-mode data arrays, the origina l data array can be replaced by a smaller coefficient array. Three-mod e principal component analysis (PCA) is then performed on the much sma ller coefficient array instead of on the original array. For the compr ession approach to be efficient the three-mode data array is assumed t o be well approximated by smooth functions. The smoothness affects the dimensions of the coefficient array. It is always possible to approxi mate the data to any precision but the reward in reduced computation t ime and storage is lost when the dimensions of the coefficient array a pproach the dimensions of the original array.