This is a fullstack machine learning project consisting of:
-
Machine Learning Model - This is in the notebooks folder. I used a pre-existing dataset from Kaggle about some students'marks in relation to many factors including age, gender, test 1 and most interestingly, alcohol consumption. I trained a model though sklearn's linear regression algorithm to estimate their final exam mark based on severalfactors. This model was then saved in a .pickle file to use in the backend Flask server.
-
Flask API - A python backend server app with an API route that accepts an array input then uses a pre-defned prediction model to return a single final exam mark.
-
React Client Web App - This client consumes the Flask API by sending it some inputs from the user then receiving an estimate of that student's final mark according to the parameters supplied.
I wish to produce the same kind of application using real data from a local university. Such a platform could be used by school counsellors and academic advisors to predict a student's final mark and offer counselling or help before it is too late. The only challenge is to collect the necessary data.