Incrementally Learning Time-Varying Half-Planes
Abstract
We present a distribution-free model for incremental learning when concepts vary with time. Concepts are caused to change by an adversary while an incremental learning algorithm attempts to track the changing concepts by minimizing the error between the current target concept and the hypothesis. For a single half(cid:173) plane and the intersection of two half-planes, we show that the average mistake rate depends on the maximum rate at which an adversary can modify the concept. These theoretical predictions are verified with simulations of several learning algorithms including back propagation.
Cite
Text
Kuh et al. "Incrementally Learning Time-Varying Half-Planes." Neural Information Processing Systems, 1991.Markdown
[Kuh et al. "Incrementally Learning Time-Varying Half-Planes." Neural Information Processing Systems, 1991.](https://mlanthology.org/neurips/1991/kuh1991neurips-incrementally/)BibTeX
@inproceedings{kuh1991neurips-incrementally,
title = {{Incrementally Learning Time-Varying Half-Planes}},
author = {Kuh, Anthony and Petsche, Thomas and Rivest, Ronald L.},
booktitle = {Neural Information Processing Systems},
year = {1991},
pages = {920-927},
url = {https://mlanthology.org/neurips/1991/kuh1991neurips-incrementally/}
}