Question about the reproducible results?

wuyifan18 / DeepLog

Pytorch Implementation of DeepLog.

MIT License

361 stars 154 forks source link

Question about the reproducible results? #32

Closed d0ng1ee closed 4 years ago

d0ng1ee commented 4 years ago

I notice your repo used the same parameters as the origin paper,but after the experiment , the result
F1-measure of this repo is only about ~89%, which is not as good as the 96% in the paper. What's your experiment result @wuyifan18 ? Thank you!

wuyifan18 commented 4 years ago

@donglee-afar just look at these codes bellow, you'll get the answer. https://github.com/wuyifan18/DeepLog/blob/a82c3128849428c252750ec783cb4bd287a396b5/LogKeyModel_predict.py#L19-L22

Yifan Wu

d0ng1ee commented 4 years ago

Amazing！I haven't noticed there is so many same sessions in the dataset. Thank you, PKU Wu Yifan!