Rainfall Prediction System

Overview

This project is a Machine Learning based Rainfall Prediction System developed using Python in Google Colab.

The system analyzes weather-related parameters and predicts whether rainfall will occur or not using a Random Forest Classifier model.

---

Live Demo

https://your-streamlit-link.streamlit.app

---

Technologies Used

• Python
• Google Colab
• Pandas
• NumPy
• Scikit-learn
• Matplotlib

---

Machine Learning Concepts Used

• Data Preprocessing
• Handling Missing Values
• Feature Selection
• Oversampling for Data Balancing
• Random Forest Classification
• Model Evaluation
• ROC Curve & AUC Score

---

Project Workflow

1. Import Libraries
2. Load Dataset
3. Clean Dataset
4. Handle Missing Values
5. Convert Target Variable
6. Balance Dataset using Oversampling
7. Feature Selection
8. Split Training & Testing Data
9. Train Random Forest Model
10. Predict Rainfall
11. Evaluate Accuracy
12. Generate ROC Curve and AUC Score

---

Features

• Predicts rainfall occurrence
• Handles missing data efficiently
• Uses balanced dataset for better accuracy
• Evaluates performance using ROC Curve
• Supports prediction on random/custom input data

---

Files Included

• MLE_PROJECT_ON_RAINFALL_PREDICTION.ipynb → Main project notebook
• README.md → Project documentation

---

How to Run the Project

1. Open the notebook in Google Colab
2. Upload the dataset file
3. Install required libraries if needed
4. Run all notebook cells sequentially

---

Model Used

Random Forest Classifier

---

Project Outputs

Performance Evaluation

AUC Score

Future Improvements

• Deploy as a web application
• Add real-time weather API integration
• Improve prediction accuracy
• Add graphical user interface

---

Author

Harshita Singh

B.Tech CSE (AI & DS) Student at Marwadi University
IITM BS Degree Student
Interested in Machine Learning and Data Science