PREDICTION OF EMPLOYEE SATISFACTION LEVEL USING OPTIMIZED NEURAL NETWORK

With the increase in global business activities, many companies expanding their business to overseas markets, Human Resource Management (HRM) is needed to make sure that they hire and retain well-performed employees. For a long time, companies/organizations have had big problems in getting accurate professionals to do the right work and training. The aim of this study is to design an automatic job satisfaction system using an optimized neural network approach. Initially, preprocessing is applied to the data to convert string data in terms of numeric data for fast computation purposes, and we change the name of "sales" by "Name of Department and "salary" by "low, medium, and high." The data analysis is performed based on three different factors, such as the number of employees in each department, the number of employees according to the salary range (low, medium, and high), and the number of employees according to the salary range and department. After this, we find out essential features (Satisfaction level and Last evaluation, Number of projects, Average monthly hours, Older employees with more than 10 years in the company), and then determine the correlation between these factors. Now, the Genetic Algorithm (GA) is applied as an optimization approach to enhance the quality of features. These optimal features are fed as input data to Artificial Neural Network (ANN), which is used for the prediction of employee's satisfaction level. At last, to show the effectiveness of the proposed work, a comparison between proposed GA with the ANN approach with the traditional GA with the K-means approach has been presented.