Weather Predictor
We will be creating a website/webapp by integrating ML algorithms which analyses the past data of weather across India and predicts the future weather. Furthermore, this model can also predict humidity, cloud patterns an
Repository Video πΊοΈWeather Predictor
π Brief Overview
A simple web application to get current weather of the city which is input by the user. The API incorporated(OpenWeatherMap API) in this web app collects the data from relevant database and fetches the information.
π‘ Key Insights
- This model takes in data such as temperature, precipitation and date
- On analysis, the machine learning model can identify patterns and trends that can help it make accurate predictions about future weather conditions
- We will be using a prop dataset from Kaggle with historical Weather data of various datasets which comes from a variety of Sources, such as weather stations, satellites, and radar systems
π― Outcomes
Once the model has been trained, it can be used to make predictions about future weather conditions based on current weather data. These predictions can be used by meteorologists, emergency responders, and other professionals to plan for and respond to weather events.
Follow this url to visit our web app:
https://prithvi2310-weather-predictor-app-3sudp2.streamlit.app
π Timeline
- This project was initiated with a blank repository on 04/03/2023
- By the end of 04/03/2023 we had small web app running and an idea to implement LSTM with it
- In the morning of 05/03/2023 we had a LSTM model using keras library but we couldn't integrate it with the web app
- The model is fully functional with an accuracy of 80%
- A simple working web app was created by the afternoon of 05/03/2023 with the functionality of fetching the live weather data of the particular city
π₯³ What worked
- The app was able to fetch data, more or less, accurately
- The UI was built successfully using the streamlit package
π’ What didn't work
- We couldn't find an API that could help us fetch historical weather data
- We decided to get an appropriate dataset (csv file) for our model but it wasn't publically available either
- We manually designed the sample.csv and the model was trained but its integration with the streamlit package failed and the desired product could not be completed
π« Contributors
Contributions make the open source community such an amazing place to learn, inspire, and create.
The contributors to this project are
πͺͺ License
The MIT License - Prithvi Shah. Please have a look at the LICENSE.md for more details.
Project created by Prithvi Shah