Developing intelligent hybrid DNN model for predicting students’ employability – A Machine Learning approach

Abstract

The employability of students is an important concern for colleges nowadays, and forecasting it can assist institutions take prompt action to improve the institutional placement ratio. The most effective data mining methods for forecasting students' employability are categorization methods. Students can concentrate on areas where they need to improve to better match the company's skill set by being aware of their deficiencies prior to an interview. Also, forecasting student employability might assist academic staff in developing curricular plans. In order for kids, parents, guardians, organisations, and businesses to profit to some extent, this report offered a fresh, futuristic roadmap. Out of seven trials, the first five were carried out using in-depth statistical calculations, while the latter two were carried out using supervised machine learning techniques. On just one extreme, the Support Vector Machine (SVM) had the highest accuracy rate of 93% when it related to forecasting employment status. The Random Forest (RF), on the other hand, was able to identify the gender of placed kids with a maximum accuracy of 89%. It is also advised to determine the placement of gender and placement status using a number of key criteria. A statistical t-test with a significance threshold of 0.05 demonstrated that the student’s gender had no impact on the pay that was provided during job placement.