NORMA eResearch @NCI Library

Evaluation of Employee Attrition by Effective Feature Selection using Hybrid Model of Ensemble Methods

Jain, Divyang (2017) Evaluation of Employee Attrition by Effective Feature Selection using Hybrid Model of Ensemble Methods. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
PDF (Master of Science)
Download (1MB) | Preview


Employees are leaving organisation's pre-maturely that results in high losses for the organisation which cannot be predicted by HR.This paper contributes to HR predictive analytics (HRPA) that helps in a predicting the employees who will leave the organisation in a certain period of time by using a hybrid model of machine learning techniques. Employee attrition is a major concern today which is related to customer attrition prediction and much research has been done on customer churn by using ensemble methods.This research explains how predicted accuracy, sensitivity and specificity can be enhanced by the use of ensemble methods in determining employee attrition with the feature selection method using efficient feature engineering, data wrangling, visualizing and analyzing results from previous models to increase the accuracy. This work has the potential for greater accuracy to improve employee retention and reducing Human Resource costs.The CRISP-DM method using three different ensemble methods was used (1-stacking) GLM, SVM, Decision Trees, KNN, (2-bagging) Random Forest and (3-boosting) GBM, Adaptive boosting (ADA). This achieved 88.85% accuracy by these techniques from which HR can place a sound strategy to raise employee retention.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
H Social Sciences > HD Industries. Land use. Labor > HD28 Management. Industrial Management > Human Resource Management > Employee Retention
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Caoimhe Ní Mhaicín
Date Deposited: 28 Aug 2018 11:31
Last Modified: 28 Aug 2018 11:31

Actions (login required)

View Item View Item