An end‑to‑end machine learning project to forecast Ethereum prices using historical data, regression models, feature engineering, and detailed visualizations.
This project focuses on exploring, cleaning, and forecasting Ethereum
(ETH) prices using multiple regression models.
Key features include:
- Data preprocessing & validation\
- Feature engineering\
- Model training (Linear Regression, Random Forest, XGBoost)\
- Hyperparameter optimization\
- Visual prediction plots\
- Model evaluation (R², RMSE, MAPE)
ethereum-forecast/
│── eth_dataset/
│ ├── Ethereum_Merged_Data.csv
│ └── ETH_8_Years.json
│
│── src/
│ └── ethereum_forecast.py
│
│── requirements.txt
│── run.sh
│── setup.sh
│── README.md
Shows how close the predictions are to real ETH prices.
Displays predicted trend lines compared to actual historical prices.
Highlights a smaller part of the dataset to show model accuracy.
- Removing duplicates\
- Handling missing values\
- Outlier treatment using IQR\
- Feature scaling (MinMaxScaler / StandardScaler)\
- Converting date to useful features (Day, Month, Year, DayOfWeek)
Model Description
Linear Regression Baseline simple model Random Forest Regressor Handles nonlinear patterns XGBoost Regressor Best-performing model
Metric Meaning
R² Score Measures model fit RMSE Measures prediction error MAPE Percentage error
git clone https://github.com/YOUR_USERNAME/ethereum-forecast.git
cd ethereum-forecast
python -m venv venv
venv\Scripts�ctivate # Windows
source venv/bin/activate # Linux/Mac
pip install --upgrade pip
pip install -r requirements.txt
python src/ethereum_forecast.py
The script generates:
- Multiple prediction plots\
- Model evaluation metrics\
- Saved visualizations in
/plots/folder
- Team of 10 Members
Each member handled different parts such as: Data Cleaning, EDA, Model Training, Visualization, Optimization, Documentation, Presentation, Dataset Research, Testing, Deployment Setup.
MIT License
If you like this project, give it a ⭐ on GitHub!