PREDICTIVE ANALYTICS TO IMPROVE THE RECRUITMENT PROCESS IN A DIGITAL TELECOMMUNICATIONS COMPANY

• Main Author: Putu Saraswati D. J., Luh
• Format: Theses
• Language: Indonesia
• Subjects: Manajemen umum
• Online Access: https://digilib.itb.ac.id/gdl/view/64695
• Institution: Institut Teknologi Bandung

The era of Industry 4.0 refers to the growing trend towards greater automation and data exchange in technologies, such as Big Data and Artificial Intelligence. The problems faced by Human Capital Management PT. XYZ to the recruitment process which takes a long time and costs a lot. This is because there is no system screening after candidate registration. So that the manual process is considered ineffective in this era. HR analytics systems can be used for a variety of purposes, including predictive analytics. Predictive analytic is the process of forecasting future learning using advanced methodologies such as machines. Predictive analytic can help in determining system problems and their remedies. To build the predictive model, the authors carried out four stages, namely data collection, data preprocessing, model building, and evaluation of the model results. The algorithm used in this method is the classification of Random Forest and Naïve Bayes. Both of these algorithms succeeded in predicting more data sets correctly, with 70% accuracy and precision, and recall above 80%. When compared to the two algorithms, Random Forest is the best for this predictive model with a higher evaluation result than Naïve Bayes. On the problems faced by PT. XYZ, predictive analytic can be an alternative solution to be used as a scoring system because it can predict the criteria that can pass to the recruitment of PT. XYZ. The random forest classifier model will be more suitable for data collection owned by PT. XYZ because it has the best accuracy, and has good recall so that this model can be used for data sets that have never been trained before.