Lightweight conditional model extrapolation for streaming data under class-prior shift

Tomaszewska P, Lampert C. 2022. Lightweight conditional model extrapolation for streaming data under class-prior shift. 26th International Conference on Pattern Recognition. ICPR: International Conference on Pattern Recognition vol. 2022, 2128–2134.

Download (ext.)

Conference Paper | Published | English

Scopus indexed
Author
Tomaszewska, Paulina; Lampert , ChristophISTA

Corresponding author has ISTA affiliation

Department
Abstract
We introduce LIMES, a new method for learning with non-stationary streaming data, inspired by the recent success of meta-learning. The main idea is not to attempt to learn a single classifier that would have to work well across all occurring data distributions, nor many separate classifiers, but to exploit a hybrid strategy: we learn a single set of model parameters from which a specific classifier for any specific data distribution is derived via classifier adaptation. Assuming a multiclass classification setting with class-prior shift, the adaptation step can be performed analytically with only the classifier’s bias terms being affected. Another contribution of our work is an extrapolation step that predicts suitable adaptation parameters for future time steps based on the previous data. In combination, we obtain a lightweight procedure for learning from streaming data with varying class distribution that adds no trainable parameters and almost no memory or computational overhead compared to training a single model. Experiments on a set of exemplary tasks using Twitter data show that LIMES achieves higher accuracy than alternative approaches, especially with respect to the relevant real-world metric of lowest within-day accuracy.
Publishing Year
Date Published
2022-11-29
Proceedings Title
26th International Conference on Pattern Recognition
Publisher
Institute of Electrical and Electronics Engineers
Volume
2022
Page
2128-2134
Conference
ICPR: International Conference on Pattern Recognition
Conference Location
Montreal, Canada
Conference Date
2022-08-21 – 2022-08-25
eISSN
IST-REx-ID

Cite this

Tomaszewska P, Lampert C. Lightweight conditional model extrapolation for streaming data under class-prior shift. In: 26th International Conference on Pattern Recognition. Vol 2022. Institute of Electrical and Electronics Engineers; 2022:2128-2134. doi:10.1109/icpr56361.2022.9956195
Tomaszewska, P., & Lampert, C. (2022). Lightweight conditional model extrapolation for streaming data under class-prior shift. In 26th International Conference on Pattern Recognition (Vol. 2022, pp. 2128–2134). Montreal, Canada: Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/icpr56361.2022.9956195
Tomaszewska, Paulina, and Christoph Lampert. “Lightweight Conditional Model Extrapolation for Streaming Data under Class-Prior Shift.” In 26th International Conference on Pattern Recognition, 2022:2128–34. Institute of Electrical and Electronics Engineers, 2022. https://doi.org/10.1109/icpr56361.2022.9956195.
P. Tomaszewska and C. Lampert, “Lightweight conditional model extrapolation for streaming data under class-prior shift,” in 26th International Conference on Pattern Recognition, Montreal, Canada, 2022, vol. 2022, pp. 2128–2134.
Tomaszewska P, Lampert C. 2022. Lightweight conditional model extrapolation for streaming data under class-prior shift. 26th International Conference on Pattern Recognition. ICPR: International Conference on Pattern Recognition vol. 2022, 2128–2134.
Tomaszewska, Paulina, and Christoph Lampert. “Lightweight Conditional Model Extrapolation for Streaming Data under Class-Prior Shift.” 26th International Conference on Pattern Recognition, vol. 2022, Institute of Electrical and Electronics Engineers, 2022, pp. 2128–34, doi:10.1109/icpr56361.2022.9956195.
All files available under the following license(s):
Copyright Statement:
This Item is protected by copyright and/or related rights. [...]

Link(s) to Main File(s)
Access Level
OA Open Access

Export

Marked Publications

Open Data ISTA Research Explorer

Web of Science

View record in Web of Science®

Sources

arXiv 2206.05181

Search this title in

Google Scholar