This project is an end-to-end Bitcoin market analytics and forecasting application developed as part of a Data Analytics Mock Internship at Mesa Community College. It demonstrates the complete data analytics lifecycle — from data ingestion and cleaning to visualization and basic predictive modeling — using Python and an interactive dashboard.
The core objective is to analyze historical Bitcoin market data (OHLCV), communicate insights through clear visualizations for non-technical stakeholders, and provide simple baseline forecasting models for short-term price trends.
Follow the steps below to run the project locally and launch the interactive dashboard.
Clone the Repository:
- git clone https://github.com/Mun-Min/Bitcoin_Forecasting.git
- cd <your-project-folder>
Install core libraries:
- pip install pandas numpy scikit-learn matplotlib seaborn plotly streamlit
Run the dashboard:
- streamlit run dashboard_app.py
🎈 Click to run application using Streamlit cloud! 🎈
- Automated data pipeline for loading, cleaning, and resampling Bitcoin market data
- Cleaned hourly and daily OHLCV datasets
- Exploratory Data Analysis (EDA) using Jupyter Notebooks
- Interactive Streamlit dashboard with:
- Candlestick price charts
- BTC trading volume and estimated USD volume
- Forecasting using Naive and Simple Moving Average (SMA) models
- Kaggle Dataset | Bitcoin Historical Data
This project intentionally uses simple baseline models:
- Naive Forecast – assumes future prices follow the most recent observed value
- Simple Moving Average (SMA) – smooths price noise and captures short-term trends
These models are transparent, interpretable, and suitable for educational analysis of financial time series without overfitting.
Click to read more about the models
├── Datasets/
│ ├── clean_btcusd_daily.csv
│ └── clean_btcusd_hourly.csv
│
├── BTC_Data_Pipeline.py
├── dashboard_app.py
├── Data_Cleaning_Notebook.ipynb
├── EDA_Notebook.ipynb
└── README.md