Bayesian Optimization for Distributionally Robust Chance-constrained Problem
Authors: Yu Inatsu, Shion Takeno, Masayuki Karasuyama, Ichiro Takeuchi
ICML 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | confirm the usefulness of the proposed method through numerical experiments. ... In this section, we confirm the performance of the proposed method in simulator and uncontrollable settings using synthetic functions and real-world simulations. |
| Researcher Affiliation | Academia | 1Department of Computer Science, Nagoya Institute of Technology, Aichi, Japan 2RIKEN Center for Advanced Intelligence Project, Tokyo, Japan. |
| Pseudocode | Yes | Algorithm 1 DRCC-BO: BO for DRCC problem |
| Open Source Code | No | The paper does not contain any statements about releasing open-source code or provide links to a code repository. |
| Open Datasets | No | The paper primarily uses synthetic functions and a SIR model simulation. These are defined and generated within the paper's context rather than being external, publicly available datasets requiring a link or specific citation for access. |
| Dataset Splits | No | The paper operates within a Bayesian Optimization framework, iteratively selecting points for black-box functions, rather than employing traditional train/validation/test splits of a static dataset. No specific percentages or counts for dataset splits are provided. |
| Hardware Specification | No | The paper does not specify any details regarding the hardware used for running the experiments (e.g., CPU, GPU models, memory, or cloud computing instances). |
| Software Dependencies | No | The paper mentions using Gaussian Process (GP) models and a Bayesian Optimization (BO) framework, but it does not list any specific software libraries, packages, or their version numbers that would be required for replication. |
| Experiment Setup | Yes | In this experiment, both design and environment variables were set to one dimension, and the following Gaussian kernels were used as the kernel functions: k(f)((x, w), (x', w')) = σ²f,ker exp(-θ-θ'² /Lf), k(g)((x, w), (x', w')) = σ²g,ker exp(-θ-θ'² /Lg), where θ = (x, w). We used the L1-norm as the distance between distributions, and set ϵt = 0.15. Here, for simplicity, we set the overestimation parameter η to 0 and the accuracy parameter to ξ = 10^-12. ... Table 2. Experimental parameters for each setting in the synthetic function. ... Table 3. Experimental parameters for each setting in the SIR model simulation. |