![]() |
PRINT ISSN : 2319-7692
Online ISSN : 2319-7706 Issues : 12 per year Publisher : Excellent Publishers Email : editorijcmas@gmail.com / submit@ijcmas.com Editor-in-chief: Dr.M.Prakash Index Copernicus ICV 2018: 95.39 NAAS RATING 2020: 5.38 |
Variable selection in discriminant analysis may be used to identify those variables which are most relevant for use in allocating future observation. It is also expected to reduce the cost of experimentation and conditional error rate by increasing the ratio of the training sample size to the dimension. Thus, feature Selection has become important task in classification and discriminant analysis. Three variable selection methods (Univariate t-test, Wilk’s lambda Criterion and Random Forests Algorithm) were used and compared in the present study for classification and discrimination to find important characters of Indian mustard. Secondary data set on 310 genotypes of Indian mustard recorded for 12 characters was used for discrimination between populations of low and high oil content genotypes of Indian mustard. Performance of the methods was assessed in terms of leave one out cross-validation error and out of bag error rate for classification. The important variables for discrimination which significantly affected the oil content were siliqua length, Secondary branches, primary branches and days to maturity with least error rate of 33.90 per cent.
![]() |
![]() |
![]() |
![]() |
![]() |