Safe Online Convex Optimization with Unknown Linear Safety Constraints
Authors: Sapana Chaudhary, Dileep Kalathil6175-6182
AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section, we analyze the performance of our SO-PGD algorithm through experiments in two different settings. |
| Researcher Affiliation | Academia | Sapana Chaudhary 1, Dileep Kalathil 1 1 Department of Electrical and Computer Engineering, Texas A&M University, USA sapanac@tamu.edu, dileep.kalathil@tamu.edu |
| Pseudocode | Yes | Algorithm 1: SO-PGD Algorithm |
| Open Source Code | No | The paper states that proofs are available in an online supplement, which references the same arXiv paper itself (Chaudhary and Kalathil 2021. ar Xiv preprint ar Xiv:2111.07430). There is no explicit statement about releasing source code or a link to a code repository. |
| Open Datasets | No | The paper describes how the data for the simulation is generated based on specific functions and noise characteristics (e.g., 'We consider a closed polytope...', 'We consider two sequences of functions...', 'The constraint noise sequence wts are i.i.d. Gaussian'). It does not use a pre-existing publicly available dataset, nor does it provide access information for its generated data. |
| Dataset Splits | No | The paper describes parameters for its algorithm's phases (exploration vs. online gradient descent using T0 and T) and noise characteristics, but it does not specify explicit training, validation, or test dataset splits in the conventional sense for evaluating a model's performance on separate data partitions. |
| Hardware Specification | No | The paper does not provide any specific details regarding the hardware used to run the experiments, such as GPU/CPU models or memory specifications. |
| Software Dependencies | No | The paper does not specify any software dependencies (libraries, frameworks, or solvers) with version numbers. |
| Experiment Setup | Yes | Experiment Setting: We consider a closed polytope... The constraint noise sequence wts are i.i.d. Gaussian with zero mean and covariance matrix 10 3I... We choose λ = 0.5 and δ = 10 3. Exploration noise is ζts are generated according to a standard Gaussian distribution and then normalized. The safe baseline action is selected randomly from the set X s. We run the experiments with T = 106 and T0 = T 2/3 = 104. |