# Combining Statistical Features and Deep Encodings for Rehearsal-Based Class-Incremental Time Series Classification

arXiv: https://arxiv.org/abs/2606.03292 · HTML: https://mlsanity.com/paper/2606.03292/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 12 of its 23 references that are in the map.
- First posted on arXiv 2 Jun 2026.

## Builds on (most cited first, 12 of 12)

- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [A decoder-only foundation model for time-series forecasting](https://mlsanity.com/paper/2310.10688/) (2023)
- [MOMENT: A Family of Open Time-series Foundation Models](https://mlsanity.com/paper/2402.03885/) (2024)
- [Open Access Dataset for Electromyography based Multi-code Biometric Authentication](https://mlsanity.com/paper/2201.01051/) (2022)
- [Class-incremental Learning for Time Series: Benchmark and Evaluation](https://mlsanity.com/paper/2402.12035/) (2024)
- [Prediction error-driven memory consolidation for continual learning. On the case of adaptive greenhouse models](https://mlsanity.com/paper/2006.12616/) (2020)
- [Overcoming Catastrophic Forgetting in Tabular Data Classification: A Pseudorehearsal-based approach](https://mlsanity.com/paper/2407.09039/) (2024)
- [Streaming Traffic Flow Prediction Based on Continuous Reinforcement Learning](https://mlsanity.com/paper/2212.12767/) (2022)
- [Decentralised and collaborative machine learning framework for IoT](https://mlsanity.com/paper/2312.12190/) (2023)
- [CEL: A Continual Learning Model for Disease Outbreak Prediction by Leveraging Domain Adaptation via Elastic Weight Consolidation](https://mlsanity.com/paper/2401.08940/) (2024)
- [TiCT: A Synthetically Pre-Trained Foundation Model for Time Series Classification](https://mlsanity.com/paper/2511.19694/) (2025)
- [PTMs-TSCIL Pre-Trained Models Based Class-Incremental Learning](https://mlsanity.com/paper/2503.07153/) (2025)

## Abstract (arXiv)

Many systems used in real-world environments require adding new categories and incorporating new information without forgetting what was previously learnt by the classification model. This is known as class-incremental continual learning, and in the case of multivariate time-series, is further complicated by the temporal structure of the data. In this paper, we present a novel approach for performing class incremental continual learning for the classification of multivariate time series data based upon the construction of a dual-stream feature extraction pipeline (using both deep temporal embedding features generated via a pre-trained frozen foundation model and application of statistical features). Evaluated on five benchmark datasets, the proposed system achieves competitive average accuracy across all datasets while maintaining low forgetting rates across all experimental configurations.