The deployed web-app is live now - Give Web-App a Try

• Developed an app that forecasts flight ticket prices with 94.88 % accuracy
• Preprocessed the data to a specific data form which is suitable for machine learning model
• Applied Dummy Encoding to Categorical Features such as Airline, Source and Destination
• Generated Feature Importance using Random Forest Algorithm
• Built a web-app using Streamlit and deployed it to the Heroku: A Cloud Application Platform

– Feature Importance Plot

– Prediction vs Actual

Link to Dataset

• Regression
• Streamlit
• Heroku