Back to projects

Data & AI
Cryptocurrency Forecasting
Time-Series Prediction with FbProphet
PythonFbProphetPandasPlotlyJupyterTime Series

About this project
This project builds an end-to-end forecasting pipeline for cryptocurrency price prediction using Facebook's Prophet library. Historical OHLCV data is fetched, cleaned, and fed into Prophet, which automatically handles seasonality, trend changepoints, and holiday effects.
The pipeline produces forward-looking price projections with confidence intervals and decomposes the signal into trend, weekly, and yearly components — giving analysts a clear picture of what's driving price movement. Results are visualised in interactive Plotly charts within a Jupyter notebook.
Key highlights
- Automated data pipeline from API fetch to forecast output
- Trend and seasonality decomposition with Prophet
- Interactive Plotly visualisations with confidence bands
- Modular design — easily adaptable to any time-series asset
Gallery
