合成数据集

合成数据集是由伦敦城市大学和华威大学的研究团队开发，旨在评估和优化固定收益领域的资产配置方法。该数据集通过改进的CorrGAN模型生成合成相关性矩阵，并利用编码器-解码器模型生成资产的其他属性，如波动性、预期回报和未来回报。数据集的创建过程涉及高级的生成对抗网络技术，旨在减少对历史数据的依赖，提高资产配置方法在样本外期间的表现。该数据集特别适用于模拟基于资产配置策略的深入分析，帮助投资者构建更稳健的投资组合。

This synthetic dataset was developed by a research team from City, University of London and the University of Warwick, aiming to evaluate and optimize asset allocation methods in the fixed income domain. It generates synthetic correlation matrices via an improved CorrGAN model, and leverages encoder-decoder models to produce other asset attributes such as volatility, expected return and future return. The dataset creation process involves advanced generative adversarial network (GAN) technologies, with the goal of reducing reliance on historical data and improving the performance of asset allocation methods during out-of-sample periods. This dataset is particularly suitable for conducting in-depth analyses of asset allocation-based strategies through simulation, helping investors build more robust investment portfolios.