search

chickenbestlover / RNN-Time-series-Anomaly-Detection

RNN based Time-series Anomaly detector model implemented in Pytorch.
Apache License 2.0
1.17k stars 316 forks source link

prediction_window_size and meaning of graph #28

Open eladwass opened 5 years ago

eladwass commented 5 years ago

Hi

just to make sure, prediction_window_size has no meaning right now hasn't it?

and, can you please clarify the meaning of the blue and green part in the output graphs?

thanks ahead

chickenbestlover commented 5 years ago

prediction_window_size is the number of steps in multi-step prediction. It is needed only in anomaly detection. In the time-series prediction figure: The green part is the case when ground-truths are used as the model's inputs The blue part is the case when model output is used as the model's input