# A benchmark of data stream classification for human activity recognition on connected objects

arXiv: https://arxiv.org/abs/2008.11880 · HTML: https://mlsanity.com/paper/2008.11880/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Aug 2020.

## Built on by (most cited first, 6 of 6)

- [Reducing numerical precision preserves classification accuracy in Mondrian Forests](https://mlsanity.com/paper/2106.14340/) (2021)
- [Mondrian Forest for Data Stream Classification Under Memory Constraints](https://mlsanity.com/paper/2205.07871/) (2022)
- [Extremely Simple Streaming Forest](https://mlsanity.com/paper/2110.08483/) (2021)
- [Unsupervised explainable activity prediction in competitive Nordic Walking from experimental data](https://mlsanity.com/paper/2406.12762/) (2024)
- [On Imbalanced Regression with Hoeffding Trees](https://mlsanity.com/paper/2602.22101/) (2026)
- [Cluster-Specific Localized Drift Detection for Efficient Batch Model Adaptation under Controlled Distribution Shift](https://mlsanity.com/paper/2606.22026/) (2026)

## Abstract (arXiv)

This paper evaluates data stream classifiers from the perspective of connected devices, focusing on the use case of HAR. We measure both classification performance and resource consumption (runtime, memory, and power) of five usual stream classification algorithms, implemented in a consistent library, and applied to two real human activity datasets and to three synthetic datasets. Regarding classification performance, results show an overall superiority of the HT, the MF, and the NB classifiers over the FNN and the Micro Cluster Nearest Neighbor (MCNN) classifiers on 4 datasets out of 6, including the real ones. In addition, the HT, and to some extent MCNN, are the only classifiers that can recover from a concept drift. Overall, the three leading classifiers still perform substantially lower than an offline classifier on the real datasets. Regarding resource consumption, the HT and the MF are the most memory intensive and have the longest runtime, however, no difference in power consumption is found between classifiers. We conclude that stream learning for HAR on connected objects is challenged by two factors which could lead to interesting future work: a high memory consumption and low F1 scores overall.