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

Anomaly detection without labels #7

Open leonardobegher opened 5 years ago

leonardobegher commented 5 years ago

The anomaly detector is really awesome! I want to know, if there is a way to adapt it, to detect anomalies without labeling the data, like the unsupervised learning. Thanks!

chickenbestlover commented 5 years ago

Hi, The network is trained in an unsupervised manner; Therefore, the labels are not used for training. However, They are needed to evaluate the performance of the network. When evaluating the performance of the network, we need to calculate the accuracy of the anomaly detection using F-1 metrics which requires the truth labels.