Evaluating the likelihood of job placements — that a candidate is going to be placed in an organization post-course completion.
Challenge
The client wanted to evaluate the likelihood of job placements, that is, the possibility of a candidate getting a placement in an organization post course completion.
Our proposal was to extract data from multiple sources; create a master data file with relevant information on colleges, courses, placements, industry, macro-economic indicators, etc.; and analyze the factors to develop/update the parameter list.
Solution
As per the observed trends, we followed the given approach:
- Focused on engineering undergraduate studies and science & technology postgraduate courses.
- Considered multiple credible data sources and ended up with 2000+ features for each segment.
- Shortlisted the features list to “23 most important features” for the preliminary scorecard.
- Created the co-applicant scorecard to be used in conjunction with the applicant scorecard. It mitigates the risk of defaults during the course.
Impact
Overall, we observed ~40% reduction in bounce from the worst-to-best categories. Also, there was ~37% reduction in bounce across categories.
Basis of the scorecard, we recommended the use of differential strategies for the various score segments.
Basis of the scorecard, we recommended the use of differential strategies for the various score segments.
