vumalo/CO2-prediction-using-Arima-model

This project focuses on forecasting daily carbon dioxide (CO₂) emissions in Rivers State, Nigeria, using time series analysis techniques. The goal is to build a predictive model that can help stakeholders monitor and anticipate environmental pollution levels in different areas across the state. Carbon Emission Forecasting in Rivers State This project focuses on forecasting daily carbon dioxide (CO₂) emissions in Rivers State, Nigeria, using time series analysis techniques. The goal is to build a predictive model that can help stakeholders monitor and anticipate environmental pollution levels in different areas across the state. Project Description The dataset used contains daily CO₂ emissions data collected from various areas in Rivers State. This analysis zooms in on data from *Mile 3*, a notable location within the state, to demonstrate the modeling process. Key steps in the project include: **Data Cleaning & Preparation**: Handling missing values and filtering emissions data by area. **Stationarity Testing**: Using the Augmented Dickey-Fuller test. **Differencing and Autocorrelation Analysis**: Visualizing ACF and PACF plots. **Model Building**: Fitting ARIMA models with different parameters. **Model Evaluation**: Examining residual errors and performance metrics such as RMSE, MAE, and MAPE. **Forecasting**: Predicting future CO₂ emission levels with confidence intervals. Tools & Libraries - Python - Pandas & NumPy - Statsmodels - Matplotlib How to Use 1. Clone this repository. 2. Load the Jupyter Notebook (`.ipynb`) and ensure the dataset `dataset_emmission.csv` is in the same directory. 3. Run the notebook cells to reproduce the analysis and generate forecasts. Next Steps - Extend forecasting to multiple locations. - Integrate real-time data pipelines. - Deploy the model using a Flask dashboard or Streamlit app.