Reproducing ECG5000 with min-max scaler

[WooheonHong]

Hi,

I have been trying to reproduce your results and particularly those on the ECG dataset
Using the mix-max scaler with the norm_method causes the following errors:

python main.py --dataset ECG_data --norm_method min_max

File "main.py", line 61, in <module>
    _, normalize_statistic = train(train_data, valid_data, args, result_train_file)
File "handler.py", line 169, in train
    json.dump(normalize_statistic, f)

TypeError: Object of type ndarray is not JSON serializable

How can we solve this?

And could you tell me the parameters for reproducing metr-la dataset?
There is too much difference from the performance stated in the paper.

Thank you.

[hangzhao-microsoft]

Hi,

Thanks for your feedback, we have fixed this bug and we recommend you to use z-score for reproducing.

[razvanc92]

I ran the code with the default settings on ECG5000 (the provided data), and I get the following results: MAPE: 0.89 | MAE: 0.31 | RMSE: 0.5171 which are significantly worse when what is presented in the paper. Any suggestions on what could be the reason?

[WooheonHong]

The code was different from the paper, so I reproduced it like a paper, but I also got extremely poor performance.
Has anyone succeeded in reproducing it?

[wangyuhu]

I also found this problem. I ran other model on ECG5000, also got poor performance: MAE: 0.34 | RMSE: 0.58.
Could the author teach us, pls?

[mb-Ma]

I ran the code with the default settings on ECG5000 (the provided data), and I get the following results: MAPE: 0.89 | MAE: 0.31 | RMSE: 0.5171 which are significantly worse when what is presented in the paper. Any suggestions on what could be the reason?

I also find this issue. Do you solve it? Or do you get any responses from the authors?

[Jimmy-Liu-JL]

I ran the code with the default settings on ECG5000 (the provided data), and I get the following results: MAPE: 0.89 | MAE: 0.31 | RMSE: 0.5171 which are significantly worse when what is presented in the paper. Any suggestions on what could be the reason?

I also find this issue. Do you solve it? Or do you get any responses from the authors?

I got this promblem, too.

[Man-Yacan]

Me too.