# Multi-Factor Risk Modeling for Cryptocurrency Price Prediction **URL:** \ **Category:** Data Science\ **Created:** [August 27, 2024, 9:44am UTC](https://forum.numer.ai/t/multi-factor-risk-modeling-for-cryptocurrency-price-prediction/7685 "2024-08-27T09:44:32Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![marianacu](https://avatars.discourse-cdn.com/v4/letter/m/c5a1d2/32.png) [@marianacu](https://forum.numer.ai/u/marianacu)\ **Post date:** [August 27, 2024, 9:44am UTC](https://forum.numer.ai/t/multi-factor-risk-modeling-for-cryptocurrency-price-prediction/7685/1 "2024-08-27T09:44:32Z")
In this analysis, I developed and evaluated a multi-factor risk model to predict cryptocurrency price movements. The model incorporates a wide range of features sourced from diverse datasets, including a 100-token dataset from Numerai. The analysis focuses on exploring the impact of various financial and macroeconomic factors on cryptocurrency prices. **Data Collection:** The data used in this study was meticulously curated from several sources, including public APIs and specialized financial datasets. A notable addition is the Numerai 100-token dataset, which was retrieved as a Parquet file for efficient processing. This dataset played a crucial role in the development of the model, providing valuable insights into the factors that influence cryptocurrency prices. **Model Development:** The model development process involved the use of advanced machine learning techniques, including Random Forest and XGBoost, to build a robust predictive model. The feature importance was analyzed using methods like Partial Dependence Plots (PDPs) and a comparison between Linear Regression coefficients and Random Forest feature importances. **Results:** The final model was evaluated based on its Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R² score. The Random Forest model outperformed other models, demonstrating high accuracy in predicting cryptocurrency prices. The impact of key features such as ATR and close\_rolling\_std\_7 was explored in detail, providing insights into the factors driving price changes in the cryptocurrency market. **Conclusion:** This study contributes to the broader understanding of cryptocurrency price movements by identifying significant factors that influence prices. The model’s high accuracy and usability suggest it could be a valuable tool for both researchers and practitioners in the field of cryptocurrency trading and risk management. I welcome any feedback from the community and am open to discussions on further improvements or collaborative projects. [Multi-Factor Risk Modeling for Cryptocurrency Price Prediction Report](https://my-report.tiiny.site) ---
**Author:** ![okpo](https://avatars.discourse-cdn.com/v4/letter/o/ce73a5/32.png) [@okpo](https://forum.numer.ai/u/okpo)\ **Post date:** [August 27, 2024, 11:34am UTC](https://forum.numer.ai/t/multi-factor-risk-modeling-for-cryptocurrency-price-prediction/7685/2 "2024-08-27T11:34:14Z")
Good job and congrats. I’m trying to upload mine too ---
**Author:** ![marianacu](https://avatars.discourse-cdn.com/v4/letter/m/c5a1d2/32.png) [@marianacu](https://forum.numer.ai/u/marianacu)\ **Post date:** [August 27, 2024, 11:42am UTC](https://forum.numer.ai/t/multi-factor-risk-modeling-for-cryptocurrency-price-prediction/7685/3 "2024-08-27T11:42:56Z")
Thank you! I appreciate it. Best of luck with your upload! Looking forward to seeing your work! ---
**Author:** ![okpo](https://avatars.discourse-cdn.com/v4/letter/o/ce73a5/32.png) [@okpo](https://forum.numer.ai/u/okpo)\ **Post date:** [August 27, 2024, 11:45am UTC](https://forum.numer.ai/t/multi-factor-risk-modeling-for-cryptocurrency-price-prediction/7685/4 "2024-08-27T11:45:38Z")
thanks but I can’t seem to find the button ---
**Author:** ![datahunter](https://avatars.discourse-cdn.com/v4/letter/d/8e7dd6/32.png) [@datahunter](https://forum.numer.ai/u/datahunter)\ **Post date:** [August 27, 2024, 12:38pm UTC](https://forum.numer.ai/t/multi-factor-risk-modeling-for-cryptocurrency-price-prediction/7685/5 "2024-08-27T12:38:41Z")
Really like the analysis, good job! ---
**Author:** ![yunusgumussoy](https://avatars.discourse-cdn.com/v4/letter/y/f14d63/32.png) [@yunusgumussoy](https://forum.numer.ai/u/yunusgumussoy)\ **Post date:** [August 27, 2024, 2:36pm UTC](https://forum.numer.ai/t/multi-factor-risk-modeling-for-cryptocurrency-price-prediction/7685/6 "2024-08-27T14:36:17Z")
Great report! I like your approach to features including technical analysis indicators, and to model comparison, good luck!