Attribute-Efficient PAC Learning of Low-Degree Polynomial Threshold Functions with Nasty Noise

Authors: Shiwei Zeng, Jie Shen

ICML 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical Our main contribution is a new algorithm that runs in time (nd/ϵ)O(d) and under the Gaussian marginal distribution, PAC learns the class up to error rate ϵ with O( K4d ϵ2d log5d n) samples even when an η O(ϵd) fraction of them are corrupted by the nasty noise of Bshouty et al. (2002), possibly the strongest corruption model.
Researcher Affiliation Academia 1Department of Computer Science, Stevens Institute of Technology, Hoboken, New Jersey, USA.
Pseudocode Yes Algorithm 1 Main Algorithm: Attribute-Efficient Robust Chow Vector Estimator Algorithm 2 SPARSEFILTER
Open Source Code No The paper does not mention providing open-source code for its methodology. It refers to 'open-source code' in related works section regarding prior research, but not its own contribution.
Open Datasets No The paper is theoretical and does not describe actual experimental training or refer to specific datasets for this purpose.
Dataset Splits No The paper is theoretical and does not describe actual experimental validation splits.
Hardware Specification No The paper is theoretical and does not describe running experiments, thus no hardware specifications are mentioned.
Software Dependencies No The paper is theoretical and does not describe running experiments or specific software implementations with version numbers.
Experiment Setup No The paper is theoretical and does not describe an experimental setup with specific hyperparameters or training configurations.