Recognition methods based on the algorithm AdDel. The algorithm AdDel is meant for choosing a subset of the most informative elements from a big original set. The algorithm consists of consecutively changing procedures of addition of the most informative elements and deletion of the least informative ones. The algorithm indicates its high effectiveness for solving choice problems for informative characteristics. It allows one to find both the structure and the best number of characteristics. Moreover, it turns out that the algorithm is useful for solving other recognition and prediction problems: forming a minimal and sufficient set of precedents (support vectors), choice of the most essential variables in regression analysis. Results of the comparison of AdDel with other algorithms for solving various applied problems are given
References in zbMATH (referenced in 1 article , 1 standard article )
Showing result 1 of 1.