search

yzhao062 / pyod

A Python Library for Outlier and Anomaly Detection, Integrating Classical and Deep Learning Techniques
http://pyod.readthedocs.io
BSD 2-Clause "Simplified" License
8.48k stars 1.36k forks source link

Metrics for unsupervised learning for various algorithms #232

Open cgarciae opened 4 years ago

cgarciae commented 4 years ago

Hey!

I am wondering if there are readily available unsupervised metrics within each of the different algorithms of how well the model thinks its doing? I am looking for something like this: How To Evaluate Unsupervised Learning Models

Its possible that the models internally already calculate similar things but they are not exposed to the user.

yzhao062 commented 4 years ago

Unfortunately unsupervised OD does not have a universal loss or metric. I know works like this one for internal validation: https://core.ac.uk/download/pdf/37525751.pdf

But they are not widely used...