Early detection of academically poor performer in first year of Engineering using student’s Non-Cognitive traits data and employing Machine Learning based Classifier

Abstract

Every Individual who go through professional programmes like Engineering, Medicine, Management will have to face not only cognitive challenges but non-cognitive too.  In India apart from higher secondary school (HSC) exam. score; there are various entrance examination that respective students have to clear with adequate ranks before being admitted into any prestigious institution like IITs (Indian Institute of Technology), NIT (National Institute of Technology) or high ranking Private Institutions. This filters assures respective spectrum of students being landed in respective programmes of respective institution which indirectly means all the students in respective programmes are at par with respect to their cognitive skills but when we analyse failure rate in first year engineering students and dropout rate; it is quite high and it is major point of concern in many universities. Even though, these students in given batch are more or less in narrow band of marks variation in terms of entrance examination scores or HSC score; the question remain as why the failure rate in first year is high?

 It leads to investigation as; it is not only cognitive abilities but non-cognitive skills of individuals which contribute in the student’s success in first year. This paper is systematic study and development of machine learning based classifier for early detection of academically poor performer in a batch which can be identified and counselled to improve the failure rate in first year of engineering.