In this project, we develop a model for predicting if a student will be admitted to college or not. We use FastAPI for serving this model
This project is done as a part of Project of the Week at DataTalks.Club. It's a continuation of another project where we used streamlit for creating the UI for the model
Dataset: https://www.kaggle.com/datasets/mahwiz/school-data
Install the devependencies
pipenv install --dev
Train the model:
pipenv run python train.py
Serve it
pipenv run uvicorn predict:app --reload
Navigate to http://localhost:8000/docs to see the docs
Test request:
{
"type_school": "Academic",
"school_accreditation": "A",
"gender": "Male",
"interest": "Uncertain",
"residence": "Rural",
"parent_age": 48,
"parent_salary": 7160000,
"house_area": 71.0,
"average_grades": 88.46,
"parent_was_in_college": true
}