Pricing with Contextual Elasticity and Heteroscedastic Valuation
Authors: Jianyu Xu, Yu-Xiang Wang
ICML 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | 5. Numerical Experiments Here we conduct numerical experiments to validate the low-regret performance of our algorithm Pw P. Since we are the first to study this heteroscadestic valuation model, we do not have a baseline algorithm working for exactly the same problem. However, we can modify the RMLP2 algorithm in Javanmard & Nazerzadeh (2019) by only replacing their max-likelihood estimator (MLE) for θ with a new MLE for both θ and η . This modified RMLP-2 algorithm does not have a regret guarantee in our setting, but it may still serve as a baseline to compare with. n the following part, we will compare the cumulative regrets of our ONSPP algorithm with the (modified) RMLP-2 in the following two scenarios: 1. The linear-fractional valuation yt = x t θ +Nt 2. A fully-linear valuation yt = x t θ + x t η Nt. |
| Researcher Affiliation | Academia | 1University of California, Santa Barbara 2University of California, San Diego. |
| Pseudocode | Yes | Algorithm 1 Pricing with Perturbation (Pw P) ... Algorithm 2 Online Newton Step (ONS) |
| Open Source Code | No | The paper does not contain an explicit statement about the release of source code or a link to a code repository. |
| Open Datasets | No | We test Pw P and the modified RMLP-2 on the demand model assumed in Eq. (1) with both stochastic and adversarial {xt} sequences, respectively. Basically, we assume T = 216 d = 2, Nt N(0, σ2) with σ = 0.5, and we repeatedly run each algorithm for 20 times in each experiment setting. |
| Dataset Splits | No | The paper describes generating simulation data but does not specify traditional training, validation, or test dataset splits. |
| Hardware Specification | No | The paper does not specify any particular hardware (e.g., CPU, GPU models, or cloud resources) used for conducting the experiments. |
| Software Dependencies | No | The paper does not specify any software dependencies with version numbers (e.g., Python, PyTorch, or specific libraries). |
| Experiment Setup | Yes | Basically, we assume T = 2^16 d = 2, Nt N(0, σ^2) with σ = 0.5, and we repeatedly run each algorithm for 20 times in each experiment setting. ... We implement and test Pw P and RMLP2 on stochastic {xt} s, where xt are iid sampled from N(µx, Σx) (for µx = [10, 10, . . . , 10] and some randomly sampled Σx) and then normalized s.t. xt 2 1. ... Here we design an adversarial {xt} sequence to attack both algorithms. |