Track: Engineering Education

Abstract

This research is focused on the development and comparison of models based on ensemble machine learning algorithms to predict professional career development preferences among students five years after their graduation using the results of the University Entrepreneurial Spirit Students’ Survey 2018. The research is carried out following a classic 4-stage methodology (analysis, design, development, and validation). During the analysis, surveys are thoroughly reviewed and preprocessed. During the design, questions are grouped and combined to build 11 predictive models. Construction and validation are carried out entirely using the software WEKA. For the purposes of this investigation 1.121 surveys are considered. Initially the dataset is split up in a subset for training and test (80%) and a subset for validation (20%). The approach to predict students’ mid-term career preferences comprised the use of an ensemble scheme (stacking) composed by a logistic regression as meta-model, and a decision tree, and a support vector machine as base models. Experimental results show that half the proposed models predict correctly around 77% of the surveys. In conclusion, ensemble models can be used to predict students’ professional career development preferences. However, predictions’ accuracy depends on the attribute selection.