Sample Average Approximation for Conditional Stochastic Optimization with Dependent Data
Authors: Yafei Wang, Bo Pan, Mei Li, Jianya Lu, Lingchen Kong, Bei Jiang, Linglong Kong
ICML 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | To verify the theoretical results for SAA for CSO with dependent data and its applications in real problems, in this section, we conduct numerical experiments, including Model-Agnostic Meta-Learning (MAML) Linear Quadratic Regulator (LQR), invariant regression, and real application of risk-averse portfolio allocation. |
| Researcher Affiliation | Academia | 1Department of Mathematical and Statistical Sciences, University of Alberta, Edmonton, Canada 2Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing, China 3School of Mathematics, Statistics and Actuarial Science, University of Essex, Colchester, UK 4Department of Mathematics and Statistics, Beijing Jiaotong University, Beijing, China. |
| Pseudocode | No | The paper does not contain any structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide any statement about open-source code availability or links to a code repository for the methodology described. |
| Open Datasets | Yes | The data are collected from Feb 2014 to Feb 2022 with daily data. ... 1 http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data library.html |
| Dataset Splits | No | The paper mentions 'training samples' and 'testing data' (e.g., 'in-sample return', 'out-of-sample return') but does not specify explicit training, validation, or test dataset splits (e.g., exact percentages or sample counts) for reproducibility. |
| Hardware Specification | No | The paper does not provide specific details regarding the hardware used to run its experiments. |
| Software Dependencies | No | The paper does not list any specific software dependencies with version numbers. |
| Experiment Setup | Yes | We replicate 50 times with initial policy K0 R1 2T with K0 ij = 0.2 for all i, j. ... wt N(0, 1). ... In our experiment, we also introduce the penalty term of r(x) = 0.1 x 1 and set the parameter, λ = 0.5 and N = 24. Let K months be a window. The incremental learning is implemented in this process: we predict the total return of a window by using the latest 24 months with K = 12. |