Vicam/Unsupervised_Anomaly_Detection

Little question for markov

Opened this issue · 0 comments

Hi Vicam,

I'm thank for your code. It's very cool.

But I have a little question about your markov model.

I see you choose real_threshold for determine a sequence is anomalous or not by threshold power window_size. So could you show me how you choose this way to scale threshold? Is it the best or exist another way better?

Please let me know.

Thanks & Best Regards,
P