College Recommendation Engine

Website
API
Model

Inspiration

I've been there. After pouring countless hours into preparing for the GRE, I found myself facing the daunting task of applying to colleges. Which university would appreciate my score? Where should I spend my limited application budget? The uncertainty and stress were overwhelming. It was my personal experience that drove me to create a solution. I realized that I wasn't alone in this predicament, that students everywhere were facing similar challenges.

What it does

In our solution, We developed a machine learning model to address the complexities I personally encountered while applying to universities after taking the GRE. It integrates data from the United States Education Directory, incorporating academic parameters like SAT and ACT scores and uses that data to recommend colleges for students when they enter their marks

How we built it

We utilized a simple KNN clustering recommendation model that gets a student's SAT and ACT scores as input and uses that data to find colleges from our dataset which contains the 25th percentile admission scores of colleges in the USA.

Challenges we ran into

The main challenge was the dataset. We wanted to use complex neural networks but we only had 1000 colleges and the data points were not diverse. The main goal of the project was to personalize foreign college admissions but we couldn't do that because there is no dataset that's readily available for foreign college admissions.

Accomplishments that we're proud of

We collaborated as a team by sharing our workload. This made it possible to complete the project on time without any issues. We learned how git works and how to scrape data from the internet. The whole experience also helped us manage time effectively.

What's next for College Recommendation Engine

Finding a dataset or scraping data that's relevant to foreign college admissions and introducing more personalized features like locations, type of degree etc.,