https://github.com/edjoukou/forecasting-co2-emissions
Forecasting carbon footprint using a statistical methods (SARIMA, Holt-Winters & Baseline naive)
https://github.com/edjoukou/forecasting-co2-emissions
Last synced: over 1 year ago
JSON representation
Forecasting carbon footprint using a statistical methods (SARIMA, Holt-Winters & Baseline naive)
- Host: GitHub
- URL: https://github.com/edjoukou/forecasting-co2-emissions
- Owner: EDJOUKOU
- Created: 2025-01-29T07:43:44.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-01-29T08:07:33.000Z (over 1 year ago)
- Last Synced: 2025-01-29T09:22:24.796Z (over 1 year ago)
- Language: Jupyter Notebook
- Size: 568 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Forecasting-CO2-emissions
## Project Overview
This project aims to forecast carbon emissions using statistical methods, including SARIMA, Holt-Winters, and the Baseline Naïve Approach.
## Data Source
The dataset used for this project was collected in January 2025 and sourced from Kaggle. It is available for download at the following link: [Carbon Dioxide Emissions Dataset](https://www.kaggle.com/datasets/ucsandiego/carbon-dioxide).
## Tools
- Jupyter Notebook IDE for data analysis
## Data Cleaning and Preparation
In the initial phase, the following tasks were performed:
1. Data loading and inspection
2. Handling data types
3. Addressing missing values
4. Renaming the target variable
## Exploratory Data Analysis
1. Basic information and statistics about the dataset
2. Visualization of the target variable (CO2 emissions)
3. Splitting the data into training and testing sets
## Data Analysis (Modeling)
1. SARIMA Model
2. Triple Exponential Smoothing Model
3. Comparison of forecasts with actual values, using the last season of the training set as a reference
## Results
Both SARIMA and Triple Exponential Smoothing models demonstrated similar performance, outperforming the Naïve Baseline model, which relied on the last season of the data.