Abstract:

The competency to predict student success in a course or program generates opportunities to enhance educational outcomes to improve graduate employment. With effective performance prediction techniques, teachers can appropriate resources and instruction more precisely. Research in this area aspires to recognize features that can be used to make predictions with the help of machine learning techniques that can refine predictions and quantify aspects of student performance on employability. Moreover, research in predicting student performance on employability strives to discover interrelated features and to connect the underlying reasons why definite features work better than others. This study is to build the Technical Skills Based Employability Prediction Model (TSBEPM) using machine learning techniques. The technical skills are the scores of the students in various programming courses. The experimental work is based on the predictions obtained by various machine learning classifiers, namely Support Vector Machine, Naive Bayes, Logistic Regression, Decision Tree, Random Forest, AdaBoost, and Artificial Neural Network. To confirm all models used, the experiments were carried out using real data collected from the graduate students at the University. With the help of performance measuring parameters, different models are formulated to be used for predicting whether a student is placed or not. Random Forest gives a maximum accuracy of 70% and F1-Score of 0.85. The model is formulated to be used for predicting whether a student is placed or not.

Key words: machine learning techniques, classification model, technical skills, performance