UDC 004.93'12
COMBINED METHOD FOR CORRECTION OF INITIAL DATA IN CLASSIFICATION TASKS
P. A. Gavrilov, post-graduate student, BMSTU, Moscow; This email address is being protected from spambots. You need JavaScript enabled to view it.
K. A. Maykov, PhD (technical science), full processor, BMSTU, Moscow; This email address is being protected from spambots. You need JavaScript enabled to view it.
The combined method for inputting missing data has been proposed. The aim of this work is to investigate features and limitations of this method used for solving classification tasks with missing data. The results of comparative analysis of the developed method and a number of input methods have been presented using the algorithm of k-nearest neighbor as a classifier. The quality of a classifier is evaluated by stratified 10-fold cross-validation. The results of conducted numerical experiments showed the expediency of application of developed method for inputting missing data in the process of dealing with important classification tasks.
Key words: machine learning, classification, missing data, data preprocessing.