The most frequently used benchmark in time series classification/regression is the [UCR datasets](https://timeseriesclassification.com/), which consists of 128 time-series datasets. Both [sktime](https://www.sktime.net/en/stable/index.html) (a Tensorflow-based Python package) and [tsai](https://timeseriesai.github.io/tsai/) (a Pytorch-based Python package) have provided comprehensive solutions for data handling and model development, including the SoTA.

The most frequently used benchmark in time series classification/regression is the [UCR datasets](https://timeseriesclassification.com/), which consists of 128 time-series datasets. Both [sktime](https://www.sktime.net/en/stable/index.html) (a Tensorflow-based Python package) and [tsai](https://timeseriesai.github.io/tsai/) (a Pytorch-based Python package) have provided comprehensive solutions for data handling and model development, including the SoTA.

Research regarding continual learning with time-series data using Avalanche has already been conducted. The research, titled "[Continual Learning for Human State Monitoring](https://arxiv.org/abs/2207.00010)", uses the [WESAD](https://archive.ics.uci.edu/dataset/465/wesad) dataset which is designed for wearable stress and affect detection. The coding information can be found at the following GitHub link: [fexed/CLforHSM](https://github.com/fexed/clforhsm).

This article, ["Continual Deep Learning for Time Series Modeling"](https://www.mdpi.com/1424-8220/23/16/7167),
provides a review of the ongoing development of CL for time series data.

_Originally posted by @irosyadi in https://github.com/ContinualAI/avalanche/issues/1490#issuecomment-1706740578_