Sparsity in Partially Controllable Linear Systems

Authors: Yonathan Efroni, Sham Kakade, Akshay Krishnamurthy, Cyril Zhang

ICML 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We also corroborate these theoretical improvements over certaintyequivalent control through a simulation study. and We present a proof-of-concept empirical study, to demonstrate the end-to-end statistical advantages of leveraging sparsity in the LQR of a PC-LQ. and Figure 1 clearly shows experimental results.
Researcher Affiliation Industry 1Microsoft Research New York, NY. Correspondence to: Yonathan Efroni <jonathan.efroni@gmail.com>.
Pseudocode Yes Algorithm 1 Learning Optimal Policy of PC-LQ and Algorithm 2 Semiparametric Least Squares
Open Source Code No The paper does not explicitly state that source code for the methodology is available or provide a link to a code repository.
Open Datasets No Synthetic PC-LQ problems were generated with i.i.d. standard Gaussian entries (for all A1, A2, A3, A12, A32, B1); The paper uses synthetic data and does not provide concrete access information for a publicly available or open dataset.
Dataset Splits No The paper describes generating synthetic data and conducting 100 trials for evaluation, but does not specify train/validation/test dataset splits needed for reproduction.
Hardware Specification No The paper does not provide any specific details about the hardware used for running its experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., library names like PyTorch with version 1.9, or specific solver versions).
Experiment Setup Yes Synthetic PC-LQ problems were generated with i.i.d. standard Gaussian entries (for all A1, A2, A3, A12, A32, B1); the diagonal blocks were normalized by their top singular values so that ρ(A1) = 1, and ρ(A2) = ρ(A3) = 0.9. and ...soft-thresholded semiparametric least-squares estimator from Algorithm 1 (with ϵ = 0.1)...