climate-analysis-project

Objective

To apply core data science techniques to analyze and predict patterns in climate change using global temperature data and related indicators.

Key Features

• Dataset selection from Kaggle
• Data cleaning in Excel and Tableau Prep
• Visual exploration using matplotlib and seaborn
• Predictive modeling using Random Forest Regression
• Climate anomaly detection with Z-scores and clustering

Datasets

• Climate Change Indicators (FAO)
• Global Weather Repository
• Others evaluated but excluded: IMDb, Reddit Data is Ugly, Sleep Dataset

Technologies Used

• Python (Pandas, Seaborn, Scikit-learn, Matplotlib)
• Jupyter Notebook
• Tableau Prep & Excel (for initial cleaning)
• GitHub for version control

ML Model Performance

• Model: Random Forest Regression
• Mean Squared Error: 0.1050 (on values ranging from -1 to 3 °C)

Future Work

• Add ARIMA time-series prediction
• Integrate weather dataset more deeply
• Explore neural networks (TensorFlow)

License

MIT License - see the LICENSE file for details.