Linear Transformations for Iterative Reduction of Variables

A classification function is a multi-valued function, where the function values for only a fraction of the input combinations are defined. Many variables in such a function are redundant, and can be eliminated. Using linear transformation, we can further reduce the number of variables. The problem is to find the linear transformation that minimizes the number of variables. This paper considers iterative methods s-MIN to reduce the number of variables, by replacing a set of s variables with other set of s − 1 variables, by linear transformations. It requires memory with the size O(nk), where n is the number of input variables, and k is the number of the registered vectors. The time complexity is O(n^sk log k). We compare the performance of various linear transformations by experiments. Also, we show a method to use s-MIN as a pre-processor for other minimizer to reduce total computation time.