Open AccessChoosing a statistical method for predicting a quantitative indicator
Author(s) -
R. G. Vlasov,
Yury Korobov
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1155/1/012031
Subject(s) - computer science , machine learning , set (abstract data type) , process (computing) , sample (material) , artificial intelligence , basis (linear algebra) , statistical analysis , training set , statistical learning , data mining , statistics , mathematics , chemistry , geometry , chromatography , programming language , operating system
The aim of the study is to find a machine learning method that allows achieving the maximum accuracy of the forecast when training on a specific data set. The article provides a comparative overview of strategies for choosing the optimal machine learning method. The process of training sample preparation is considered, the necessity of its modification is justified. The search for the optimal machine learning method was carried out in the statistical computing environment “R” using the package “CARET”. The graphical results of training 96 algorithms that implement various statistical methods are presented, and the choice of the best one is justified on the basis of the proposed methodology.