search

titicaca / spark-iforest

Isolation Forest on Spark
Apache License 2.0
226 stars 90 forks source link

abount the anomaly score #1

Closed puppyapple closed 6 years ago

puppyapple commented 6 years ago

Hi, thanks for you implemention of IForest in spark. I have a little question about the anomaly score when prediction. In the original paper, the lower the score is, the more normal the sample is. But in your description you mentioned the opposite way. Just want to be sure.

titicaca commented 6 years ago

Yes, you're right. It was a mistake. I have just corrected it. Thank you for the reminder!