Machine Learning Web Application - California Housing Prices
Project status - Complete
A web application which deploys a Machine Learning model into production to make predictions of housing prices. The user can adjust all of the parameters that feed into the model, and then receive an updated prediction.
The app is built using Streamlit, which makes it very easy to build and deploy a data-based application such as this, and deployed to the cloud using Heroku. The model is built and trained through scikit-learn.
Please note that the model file itself is too large to be stored in this repository. The california_housing.py script contains the necessary code to create and train the model yourself (the relevant code is commented out on lines 89-91).
All feedback is warmly received. Craig Dickson can be contacted at craigdoesdata.de.