On the Statistical Consistency of Risk-Sensitive Bayesian Decision-Making

Authors: Prateek Jaiswal, Harsha Honnappa, Vinayak Rao

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

Reproducibility Variable Result LLM Response
Research Type Experimental We illustrate our theoretical findings in parametric and nonparametric settings with the help of three examples. We also present some simulation results with the single product (1-d) newsvendor problem, which is summarized here in Figure 1.
Researcher Affiliation Academia Prateek Jaiswal Department of Statistics Texas A&M University College Station, TX 77843 jaiswalp@stat.tamu.edu; Harsha Honnappa School of Industrial Engineering Purdue University West Lafayette, IN, 47906 honnappa@purdue.edu; Vinayak A. Rao Department of Statistics Purdue University West Lafayette, IN, 47907 varao@purdue.edu
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks. Methods are described in prose and mathematical formulations.
Open Source Code No The paper does not provide any explicit statements about releasing source code or links to a code repository for the methodology described.
Open Datasets No For the single-product newsvendor and Gaussian process classification problems, data is described as 'i.i.d random samples drawn from P0' or 'generated from the model', indicating simulated data, not a publicly available dataset. For the eight-schools model, it mentions 'We consider the eight-schools problem [33, 15]' which is a well-known problem setting, but it does not provide concrete access information (link, DOI, or specific citation for the dataset itself) for public availability.
Dataset Splits No The paper states for the eight-schools model: 'due to small size of the dataset in this example, the ERR is evaluated on the training data itself.' This indicates no separate validation split for training or evaluation.
Hardware Specification No The paper does not provide any specific hardware details such as GPU models, CPU types, or memory specifications used for running the experiments.
Software Dependencies No The paper does not provide specific software names along with their version numbers that are necessary for replicating the experiments.
Experiment Setup Yes For the single-product newsvendor model: 'We fix θ0 = 0.1, b = 1, h = 5, α = 1, and β = 4.1. We run RSVB algorithm with γ {0( naive ), 1, 2, 4.5, 5, 6} and repeat the experiment over 100 sample paths.' For the eight-schools model: 'We modified the experiments in [15] by introducing γ = {5, 2.5, 1, 0.5} = {0.2 1, 0.4 1, 1 1, 2 1}.'