What's the best open source (i.e., free) approach/library/tool for unsupervised/semi-supervised[i.e., with limited to no training data] time-series [like this - https://github.com/numenta/nupic/blob/master/src/nupic/datafiles/extra/nycTaxi/nycTaxi.csv] anomaly detection.
Within "R" the is the package "Anomalize" that uses both inner quartile range and generalized extreme studentized deviation to detect anomalies in time-series
R Package anomalize. See also https://cran.r-project.org/web/packages/anomalize/vignettes/anomalize_methods.html
you can use python software which is an open source and it is increasingly becoming popular among data scientist. There are many libraries available in python for both supervised and unsupervised machine learning.
https://machinelearningmastery.com/machine-learning-in-python-step-by-step/
Kind regards
Thushara
Vishal Gupta I have published a paper on anomaly detection. Hope this helpsPreprint A Research Study on Unsupervised Machine Learning Algorithms...
- Badges
- Science topic