A Comparison of Performance Prediction in The Computer Professional Groups between Multiple Discriminant Analysis and Multinomial Logistic Regression Analysis
The purpose of this study was to create the equation and compare the accuracy of predicting to be the computer profession groups (Programmer, IT Support and System Analysis) in Bangkok Thailand. There are two statistical analysis in this article: Multiple Discriminant Analysis: MDA and Multinomial Logistic Regression Analysis: MLR. The psychology questionnaire by John L. Holland is used in this article for classifying 3 aspects: Data, Personal, and Tool. 89 persons, who work in computer field, participated in the study. The result of accuracy with 3 groupsí computer professional showed that the overall accuracy is 66.7%. Furthermore, Programmer and IT Support has the overall accuracy is 93.1% and System Analyst is a high overlapping.
Keywords- Prediction, Computer Profession, Human Resource Management, Multivariate Analysis, Multiple Discriminant Analysis, and Multinomial Logistic Regression Analysis