kevin Liu

您好，非常感谢对于我们工作的关注，目前代码中默认是采用CI的方式，我最近逐步接手该项目，按照我的理解，其中对于x进行时序分解是考虑到在channel维度进行处理最好能够去除趋势特性，纯粹的季节特性可能对于channel维度处理会更加容易，目前CD代码中对于channel的处理目前还比较简单，这样可能可以更好发挥CD之间的特性。

Thank you very much for your suggestion. We will test according to your plan. Thank you again for your support of our work.

Thank you so much for your interest in our work! We truly appreciate your suggestions and are carefully considering your request. As I've recently taken over the project, and with...

您好，非常抱歉未能够及时回复，frequency_map是M4数据集中一个参数，目前freq支持的参数是options:[s:secondly, t:minutely, h:hourly, d:daily, b:business days, w:weekly, m:monthly]，还不支持30min中的选项；我们目前参考的TSLib中，exp_short_term_forecasting主要用于M4数据集预测，目前M4的评测中主要使用SMAPE作为loss，last_insample_window也是M4数据的dataloader中的一个选项，如果您不要使用可以去掉这部分代码，如果您需要使用其他数据集，可能需要对代码进行适当的改造，你可以参考目前TSLib中的最新的代码结构，由于目前我刚接手该项目，如果有不及时或者理解不到位的地方，还请您能够包含，再次感谢您对我们工作的支持。

Thank you for your interest in our work! We appreciate your thoughtful questions. For the performance comparison between TimeMixer and TimeMixer++, you can refer to the details presented in our...

Thank you very much for your suggestion. Indeed, the DFT module should be further improved. We have also received many recommendations from colleagues in the signal processing community who are...

您好，感谢对于我们工作的关注，按照我的理解，如果您需要将数据还原会原始尺度，您可以调用data loader中的inverse_transform方法，可以将数据重置回原始尺度。

您好，非常感谢您的提示，这里我们之前参考TSLib的代码上有所疏漏，我目前已经修复了，感谢您的帮助和提示，如果我目前修复存在问题的地方，还请您可以指导交流。UEA是一个时序分类的经典数据，包含了10个子数据集，具体可以参考TimesNet中对于该数据集的介绍。

Thank you for your interest in this work and for your advice! We have a number of issues to address, so it might take us some time to work through...

Thank you very much for your suggestion regarding Future Temporal Features. In my recent work, I have also found that Future Temporal Features are crucial for real-world forecasting. Unlike theoretical...