Skip to content

pradeish29/cloud-burst-predict-streamlit

Repository files navigation

Cloud Burst Prediction using Random Forest Model and Streamlit

Overview

This project aims to predict cloud burst occurrences using a Random Forest model deployed via a Streamlit web application. The model utilizes real-time weather data from OpenWeather.org, allowing users to input a city name to retrieve relevant weather information and receive a prediction output.

Deployment

this project has been deployed using streamlit cloud.

https://cloud-burst-prediction.streamlit.app/

Page view

Streamlit Page

Features

  • Real-time Weather Data: Integration with OpenWeather.org provides up-to-date weather information.
  • User-Friendly Interface: Streamlit app allows users to easily input city names and view prediction outputs.
  • Random Forest Model: Utilizes a trained Random Forest model to predict cloud burst occurrences based on weather data.

Usage

To use the application:

  1. Clone the repository to your local machine.
  2. Install the necessary dependencies using pip install streamlit joblib requests bz2file.
  3. Check and modify the path for certain files in the code
  4. Run the model.ipynb file
  5. Run the Streamlit app by executing streamlit run app.py in your terminal.
  6. Access the app via the provided local URL in your browser.
  7. Enter a city name to retrieve weather data and receive the cloud burst prediction output.

File Structure

  • app.py: Streamlit application script.
  • model.ipynb: Pre-trained Random Forest model for cloud burst prediction.
  • model.joblib: For importing and exporting the Random Forest model

Data Sources

Credits

  • Developed by: Pradeish Misara

License

This project is licensed under the MIT License - see the LICENSE.md file for details.

Feel free to customize this template further according to your project's needs!