Factors Influencing Employability of Engineering Students and Predictive Modelling with Machine Learning Algorithms

The study emphasizes the following two objectives: identification of factors influencing the employability of engineering students (placements) and investigation of how the identified factors can be used to train machine learning algorithms to predict employability. To meet the first objective, the authors carried out a Systematic Literature Review (SLR) using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) method. The SLR helped the authors not only to identify the factors influencing employability but also to explore various data sources and statistical and Machine Learning (ML) algorithms used by researchers in this direction. To address the second objective, the authors experimented with multiple ML algorithms, including Logistic Regression, Support Vector Classifier, K-Nearest Neighbors (K-NN), Random Forest, XG Boost, Voting Classifier, and Bagging, across a diverse dataset. Voting Classifier that combines XG Boost and Random Forest algorithms achieved an accuracy of $89.9 \%$, highlighting its effectiveness in predicting employability by considering age, gender, stream, internships, Cumulative Grade Point Average (CGPA), day scholar/ resident of a hostel and history of backlogs as influencing factors. The study’s outcome helps educators and policymakers create robust predictive tools that can assist in the early identification of students at risk of employment, enabling timely and effective interventions.