https://github.com/sandoche/btc-candlestick-predictions
Bitcoin candlestick predictions using lagged features and machine learning algorithms in R
https://github.com/sandoche/btc-candlestick-predictions
Last synced: 6 months ago
JSON representation
Bitcoin candlestick predictions using lagged features and machine learning algorithms in R
- Host: GitHub
- URL: https://github.com/sandoche/btc-candlestick-predictions
- Owner: sandoche
- License: mit
- Created: 2025-03-29T21:48:00.000Z (7 months ago)
- Default Branch: main
- Last Pushed: 2025-04-23T20:01:11.000Z (6 months ago)
- Last Synced: 2025-04-23T21:19:38.187Z (6 months ago)
- Language: R
- Homepage:
- Size: 5.76 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Bitcoin candlestick predictions using lagged features and machine learning algorithms in R
> Training machine learning algorithms to predict the next candlestick's direction of the bitcoin chart using a set of lagged features
## ⚙️ Dataset
Data are already downloaded and stored in the `data` folder for the following paramters:
```R
trading_pair <- "BTC-USD"
start_date <- "2024-01-01"
end_date <- "2025-03-29"
candlestick_period <- 3600
```
If you want to download the data for different parameters, you can run the `data-downloader.R` script.
## 📊 Report
- Report in RMarkdown: [REPORT.Rmd](REPORT.Rmd)
- Report in PDF: [REPORT.pdf](REPORT.pdf)
## 🖥 Analysis and weights
- Analysis script in R: [analysis.R](analysis.R)
- Weights for the models:
- Pull them from hugging face into the `models` folder
- Weights for the models: [Weights on hugging face](https://huggingface.co/Sandoche/btc-candlestick-predictions)
- Run the following command to pull the weights:
```bash
cd models
huggingface-cli download Sandoche/btc-candlestick-predictions --local-dir .
```
## 📄 License
[MIT](LICENSE)
## 📖 Citation
```
@misc{bitcoin-candlestick-predictions,
author = {Sandoche Adittane},
title = {Bitcoin candlestick predictions using lagged features and machine learning algorithms in R},
year = {2025},
url = {\url{https://github.com/sandoche/btc-candlestick-predictions}},
}
```