rob-med / awesome-ts-anomaly-detection Goto Github PK

谢啦

I want to detect time series anomaly in price return series of stocks. Which python library is best for this objective ?

I don't see a guide for contributors, so my apologies if this is not the preferred way to suggest a new item to the list.

I wanted to introduce ADTK, a new Python package that we open-sourced recently. It is a framework for building rule-based/unsupervised anomaly detection models. It focuses on pragmatic pipelines of anomaly detection for real-world problems instead of implementation of new algorithms.

I tried to create a new branch and PR but it seems I don't have the permission.

Thank you!

Easy-to-use, powerful, and unified full pipeline automated time series toolkit.

Supports forecasting, classification, regression, and anomaly detection.

Including a large collection of TS models, including statistical models (Prophet, ARIMA, VAR, iForest, etc.), deep learning models(DeepAR, GRU, LSTNet, NBeats, VAE, etc.), and neural architecture search.

Github link: https://github.com/DataCanvasIO/HyperTS

Thanks for the suggestion!
I added licenses in the latest commit 👍

Originally posted by @rob-med in #2 (comment)

I think you summed it up very well! So I translated it into Chinese in the hope that more people will learn about these great tools.

https://github.com/SharkFin-top/awesome-TS-anomaly-detection_zh

It is a folder for "Automatic extraction of relevant features from time series", have a check here. That should be included in your list

https://github.com/unit8co/darts

I think this is amazing. It have many convenient APIs of many neural network models, like TCN, N-Beats and so on. I LOVE IT!

Do you want to add darts on the list?

Hello,
It is a great initiative to bring together all timeseries frameworks in the same place.
There is no mention of audio frameworks (e.g., librosa). I can complete the list if it is required.

It would be very convenient if you also added type of license in one of the columns.