Employee attrition prediction using machine learning models | |
| Author | Babu, Sandalla Siva |
| Call Number | AIT RSPR no.IM-23-01 |
| Subject(s) | Employee retention--Data processing Personnel management--Mathematical models Machine learning |
| Note | A research study submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Information Management |
| Publisher | Asian Institute of Technology |
| Abstract | Employees are the most precious resource in the firm. Workforce productivity has a substantial impact on the competitiveness of other firms. It is critical to have stable and collaborative staff in order to build and maintain an appropriate environment. The human resource (HR) department should assist in the development of a working atmosphere by reviewing employee database records. The administration may improve decision-making and reduce worker attrition by examining this data. The main objective of the study is to examine several prediction models and different sampling procedures and then determine the best combination of sampling strategy, prediction model, and data visualization. The optimum combination discovered is then selected for deployment using an application created to assist in determining staff attrition. To handle the imbalance and reduce the bias produced by the imbalance, multiple sampling approaches such as SMOTE, BORDERLINE-SMOTE, and ADASYN are utilized in this work to resample the data. The resampled data is fed into several prediction models such as SVM, Logistic Regression, Nave Bayesian, Random Forest, and Feedforward neural network. The most accurate sampling approach and prediction model combination are then chosen and provided via an application, which aids in determining whether or not a person will leave a firm and also adds the data to the excel form interface where we get the data inputs from the users and from there I have connected the google sheet to Looker studio where we use the data inputs that we have gathered from the employees to visualize the data and conducted statistical analysis in R studio to identify the significant factors that influence the employee attrition. Flask, HTML, CSS, and Bootstrap are used to develop the application, and Flask is used to deploy the prediction model. |
| Year | 2023 |
| Type | Research Study Project Report (RSPR) |
| School | School of Engineering and Technology |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Information Management (IM) |
| Chairperson(s) | Vatcharapon Esichaikul; |
| Examination Committee(s) | Chutiporn Anutariya;Huynh, Trung Luong; |
| Scholarship Donor(s) | AIT Partial Scholarship; |
| Degree | Research studies project report (M. Eng.) - Asian Institute of Technology, 2023 |