Hundreds of thousands of undergraduate students in India apply to graduate programs in the United States every year. A single student’s application requires many components such as a GRE score, TOEFL score, Statement of Purpose, Letters of Recommendation, CGPA, Research Ex- perience, Conference Publications, Internship Experiences, Industry Experience, Research Experience, and others. Applying to a single university can cost hundreds of dollars in application fees, sending test scores, and more. This project aims to solve this problem by predicting whether an appli- cant will be admitted to a university or not. The prediction is done by analysing Edulix.com’s data of Indian students who previously applied to graduate programs in the US.
Python version: 3.7.4
All the preprocessing and modelling is contained within the files Code.py and Code.ipynb. It's recommended to open the Jupyter Notebook as the code is divided into sections and further split into cells with outputs.
The classifiers we used are:
- SVM
- Random Forest
- Logistic Regression
- K-Nearest Neighbors
- Naïve Bayes
- Multilayer Perceptron
- Neural Networks (Tensorflow)
- edulix.com.
- B. Ghai. Analysis & prediction of american graduate admissions process. Master’s thesis, Department of Computer Science, Stony Brook University, Stony Brook, NY 11794, 2018.
- joemanley201. universityrecommendationsystem. Github.com.
- R. Swaminathan, J. M. Gnanasekaran, S. Krishnakumar, and A. S. Kumar. University Recommender System for Graduate Studies in USA. PhD thesis, University of California San Diego, 2015.
- A. Waters and R. Miikkulainen. Grade: Machine learning support for graduate admissions. In Twenty-Fifth Innovative Applications of Artificial Intelligence Conference, 2013.