Feature Selection Using Typical e: Testors, Working on Dynamical Data
暂无分享,去创建一个
Typical e:testors are useful to do feature selection in supervised clas- sification problems with mixed incomplete data, where similarity function is not the total coincidence, but it is a one threshold function. In this kind of problems, modifications on the training matrix can appear very frequently. Any modifica- tion of the training matrix can change the set of all typical e:testors, so this set must be recomputed after each modification. But, complexity of algorithms for calculating all typical e:testors of a training matrix is too high. In this paper we analyze how the set of all typical e:testors changes after modifications. An al- ternative method to calculate all typical e:testors of the modified training matrix is exposed. The new method's complexity is analyzed and some experimental results are shown.
[1] José Francisco Martínez Trinidad,et al. The logical combinatorial approach to pattern recognition, an overview through selected works , 2001, Pattern Recognit..
[2] José Ruiz-Shulcloper,et al. An overview of the evolution of the concept of testor , 2001, Pattern Recognit..
[3] José Ruiz-Shulcloper,et al. Mathematical algorithms for the supervised classification based on fuzzy partial precedence , 1999 .