Rain Prediction in Australia using Machine Learning

This project predicts whether it will rain tomorrow in Australia using historical weather data and different machine learning algorithms.

Project Overview

• Dataset: Rain in Australia dataset
• Goal: Predict the binary target RainTomorrow (Yes or No)
• Techniques: Data preprocessing, feature engineering, model training, evaluation
• Tools: Python, Pandas, NumPy, Scikit-learn, Matplotlib, Seaborn

Workflow

1. Data Loading & Exploration

   • Handle missing values
   • Summary statistics & distributions
   • Visualizations of weather patterns

2. Data Preprocessing

   • Encode categorical variables
   • Feature scaling (MinMax/Standard Scaler)
   • Train-test split

3. Modeling
   Algorithms used:

   • Logistic Regression
   • K- Nearest Neighbor
   • Support Vector Machine
   • Decision Tree

4. Evaluation Metrics (Decision Tree)

   • Accuracy = 89.04 %
   • Precision, Recall, F1-score
   • ROC-AUC Curve = 0.8908

5. Results & Insights

   • Best performing model and its metrics
   • Feature importance analysis

Results

• Logistic Regression: (add results)
• Random Forest: (add results)
• XGBoost: (add results)

Future Improvements

• Hyperparameter tuning with GridSearchCV/RandomizedSearchCV
• Try deep learning models (LSTM for time-based weather data)
• Deploy with Streamlit/Flask

How to Run

1. Clone this repository:

   git clone https://github.com/your-username/Rain-Prediction-Australia-using-ML-algorithms.git
   cd Rain-Prediction-Australia-using-ML-algorithms