Pricing Cryptocurrency Options with Machine Learning Regression for Handling Market Volatility

This replication folder contains all necessary resources to reproduce the findings of our study on the application of machine learning models to price cryptocurrency options. It includes a Jupyter notebook with the complete codebase, structured to guide users through data loading, model application, and result generation. Hyperparameters optimized via grid search are provided in pickle files. The dataset includes preprocessed cryptocurrency option data from Binance necessary for replication. Due to licensing restrictions, equity option data is not included; however, guidelines on accessing this data through Wharton Research Data Services (WRDS) are provided.

本复现文件夹包含全部必要资源，可复现我们针对机器学习模型应用于加密货币期权定价的研究结论。文件夹内包含一份完整承载代码库的Jupyter笔记本（Jupyter Notebook），其架构设计可引导用户依次完成数据加载、模型应用与结果生成全流程。通过网格搜索优化得到的超参数以pickle文件形式提供。本数据集涵盖来自币安（Binance）的预处理加密货币期权数据，足以支撑研究复现工作。受授权许可限制，股票期权数据未被纳入本次数据集；不过我们提供了通过沃顿研究数据服务（Wharton Research Data Services，WRDS）获取该类数据的操作指引。