Student’s Employability Indexing Using Machine Learning Approach

Predicting unemployment rates has become crucial since it can assist governments in making decisions and designing strategies. Conventional univariate time series models and econometric approaches for predicting unemployment rates have garnered considerable attention from governments, corporations, research institutes, and academics in prior studies. Novel approaches for forecasting unemployment rates based on search engine query data have recently been developed. To help comprehend how graduates are hired, four learning algorithms were used: SVM, DT, XGBoost, RF. The accuracy, precision, and recall metrics, as well as the f1-score and support measures, were all assessed using the efficiency matrix. During the studies, the Extreme Gradient Boosting (XGBoost) achieved an accuracy rate of 81.22 percent, which was much higher than all of the learning algorithms, including SVM 79.6 percent, DT 79.4 and RF 79.1 percent. This study is quite promising, which has pushed the researchers to improve the procedure and evaluate the developed prediction model for further study.