SimTSC
SimTSC copied to clipboard
[SDM 2022] Towards Similarity-Aware Time-Series Classification
pydtw
Sorry, because I am a novice, I do not understand this python wrapper of UCR DTW command need to input where ![Uploading QQ图片20220912173339.png…]()
由于python install -e .调试了很久不能运行,可以共享一下生成的距离矩阵吗
I meet a problem about dtw . I install pydtw by `pip install pydtw==2.0.2` and success,but i run create_dtw.py has a error . 
您好,最近拜读了你们的文章,想在别的数据集上复现一下。然后我对inductive learning这部分比较感兴趣,看你们文章中意思是包含了所有train split里面没有按few shot learning标记上的数据。但是在看代码的时候,没有找到是在哪里切换supervised learning和inductive learning,所以想在这里问一下,谢谢了!
Hello! Is it possible to show the full results for 128 UCR datasets? The paper shows the average ranks but no full results (acc) for it. Thanks!
About pydw
I follow your note, but there are some trouble with me. 1 I can't install pydtw, it It prompts such errors  2 the link of "dtw" file is removed,...
作者您好,最近有幸拜读了您的论文,想在自己的工业数据上实践一下,发现有一个比较棘手的问题。 公式(3.6)是建立我们时序相似度的矩阵,矩阵的大小为N*N,如果我没理解错的话N代表的是序列的数量,那如果我想要计算一些比较大(10W条以上的数据量)的数据集的相似度矩阵的话,可能就会出现内存不足的报错,想请问一下这样的实际情况是否有一些针对性的方法呢 附上我用自己数据及创建dtw矩阵报错的信息 MemoryError: Unable to allocate 1.02 TiB for an array with shape (374147, 374147) and data type float64