daizy-jepchumba/CO2-Emissions-Prediction

Predicting CO₂ emissions for Kenya (2020–2030) using machine learning and Kaggle dataset. # SDG13 — Predicting CO₂ Emissions (Daizy Jepchumba) ## Project Overview This project predicts country-level CO₂ emissions (kt) using historical data from Kaggle. **SDG:** 13 — Climate Action ## Dataset Source: CO₂ Emissions by Country — Kaggle Path: `/kaggle/input/co2-emissions-by-country/co2_emissions_kt_by_country.csv` ## Approach - Exploratory Data Analysis (global and per-country trends) - Models: - Baseline: Linear Regression (country + year) - Improved: Random Forest Regressor - Evaluation: MAE, MSE, R², and visual actual vs predicted plots ## Results Model Evaluation: Mean Absolute Error (MAE): 1278837.16 Mean Squared Error (MSE): 6726124881346.41 R² Score: 0.00 Kenya R2: 0.7978302393535925 ## How to run 1. Open `notebook.ipynb` in Jupyter/Kaggle. 2. Run all cells. 3. See the “Model Evaluation” outputs and plots. ## Author Daizy Jepchumba Kiplagat — daisyjepchum@gmail.com