DeepADoTS
DeepADoTS copied to clipboard
KDD-OpenSource
Repository of the paper "A Systematic Evaluation of Deep Anomaly Detection Methods for Time Series".
Hello,I found a performance issue in the definition of `fit` , src/algorithms/donut.py, [tf_session = tf.Session](https://github.com/KDD-OpenSource/DeepADoTS/blob/88c38320141a1062301cc9255f3e0fc111f55e80/src/algorithms/donut.py#L162) was repeatedly called in [for col_idx in trange(len(X.columns)):](https://github.com/KDD-OpenSource/DeepADoTS/blob/88c38320141a1062301cc9255f3e0fc111f55e80/src/algorithms/donut.py#L160) and was not closed. I think it...
I have tried running the repository in various ways and none of them was successful. Can someone please care to lay a hand ?
The DAGMM model gave the option to change hidden size, but it was actually fixed in the model... in line 76 of DAGMM.py: self.hidden_size = 5 + int(X.shape[1] / 20)
When I tried to run DGAMM-LSTM on gpu I got the following message: can't convert cuda:0 device type tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory...
This looks like an excellent repo for benchmarking. Could you please include the link of the corresponding paper in the ReadMe? I am unable to find anything by the title...
I am running dagmm with Kdd - cup dataset and I'm getting negative determinant, which avoid learning ' Does someone hit this issue? Thanks
Hi, Is there any way you can use these methods for univariate data? Thanks very much!
Dear @WGierke and team, According to your reference "Outlier Detection Using Replicator Neural Networks" , DaWaK 2002, I see that in the paper, they implemented the RNN with 5 layers....
Thank you for your excellent work. Could you provide other examples similar to the sample code of mnist autoencoder, such as dagmm, Donut, LSTMAD? I tried it myself but found...
Metadata
Owner
Metadata
Repository of the paper "A Systematic Evaluation of Deep Anomaly Detection Methods for Time Series".