While solving the problems of technical diagnostics with machinery learning involvement, there is binary classification of an object state performed: the objects are subdivided into “good” and “bad” with the help of models, received as per learning samples. The quality of classification, which specifies the efficiency of machine learning, depends on several factors, such as: the scope of original sample, method of machine learning, method of dividing the sample into learning and validating parts, selection of value indicators, etc. Sometimes it is reasonable to use aggregate methods of classification, which are, in fact, the joined results of classification basic methods. To search the best aggregate method, one iterates over all possible basis sets.
Paper
Full text
The use of aggregate classifiers in technical diagnostics, based on machine learning
Semantic Scholar · Computer Science · 2017
Abstract
While solving the problems of technical diagnostics with machinery learning involvement, there is binary classification of an object state performed: the objects are subdivided into “good” and “bad” with the help of models, received as per learning samples. The quality of classification, which specifies the efficiency of machine learning, depends on several factors, such as: the scope of original sample, method of machine learning, method of dividing the sample into learning and validating parts, selection of value indicators, etc. Sometimes it is reasonable to use aggregate methods of classification, which are, in fact, the joined results of classification basic methods. To search the best aggregate method, one iterates over all possible basis sets.