Development Stack
The notebook runs without downloading the datasets (Thanks to gdown). Just click, Learn and Explore.
If you forked this repository, you can link it to your heroku app afterwards 🔗.
Summary
The model was build by using the following Dataset. Logistic Regression, Decision Trees Classifier, Random Forest Classifier, Support Vector Machine, Multinomial Naive Byaes, KNearestNeighbor and Extra Trees Classifier was used to train the model. Support Vector Machine gave us the best model metrics, so I converted it into a pickel file and saved it.
Run Deployed Model
Click on the icon above to run the single page application in your browser directly. Just throw in the message and the model will predict whether you it is a spam or ham (legitimate) message 🙅♂️.