lejohntanjunyen/banknote_authenticator

A mini project to understand the different deployment methods with Flask, flasgger, Streamlit and Heroku

Mini Project on Bank Note Authenticator

• built a machine learning model to predict whether the bank note is authentic or not
• deployed the model using Streamlit into Heroku
• https://banknote-authenticator-1006.herokuapp.com/

Files: BankNote_Authentication.csv

• bank note data set

TestFile.csv

• test data set

BankNoteAuthenticationModel.ipynb

• RandomForestClassifier Model to predict authenticity of Bank Note

model.pkl

• pickled out RandomForestClassifier Model

prediction.py

• a simple python file to deploy model into web application using flask

flask_api.py

• similar to prediction.py but included flasgger

app.py

• python file to deploy model using Streamlit

requirements.txt

• a lsit of required packages and its version

setup.sh

• shell file with shell command
• Note: do not open this file in your computer, will create a config.toml file inside the .streamlit file and running streamlit will pose problems

Procfile

• specifies the shell file to run app on Heroku