Optimizing Discount and Reputation Trade-Offs in E-Commerce Systems: Characterization and Online Learning

Authors: Hong Xie, Yongkun Li, John C. S. Lui7992-7999

AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conduct experiments on a dataset from e Bay to show that our QLFP algorithm improves the profit by as high as 50% over both the classical Q-learning and the speedy Q-learning algorithm. Our QLFP algorithm also improves the profit by as high as four times over the case of not providing any price discount.
Researcher Affiliation Academia Hong Xie,1,2 Yongkun Li,3 John C.S. Lui2 1College of Computer Science, Chongqing University, China 2Department of Computer Science and Engineering, The Chinese University of Hong Kong 3School of Computer Science and Technology, University of Science and Technology of China hongx87@gmail.com, ykli@ustc.edu.cn, cslui@cse.cuhk.edu.hk
Pseudocode Yes Algorithm 1 : Discount Selection Via Q-learning
Open Source Code No The paper cites a technical report available at 'https://1drv.ms/b/s! Akq QNKu LPUb Edg LEi MLJQu8Mf ZM' but does not explicitly state that the source code for the methodology is openly released or provide a link to a code repository.
Open Datasets Yes We use a dataset from e Bay (Xie and Lui 2017), which contains 19,217,083 transactions of 4,586 sellers.
Dataset Splits No The paper does not provide specific training/test/validation dataset splits. It describes its experimental evaluation on an e Bay dataset but does not detail how it was partitioned for training, validation, or testing.
Hardware Specification No The paper does not provide any specific hardware details (e.g., GPU/CPU models, memory, or cloud resources) used for conducting the experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., programming languages, libraries, or solvers with their versions) used in the experiments.
Experiment Setup Yes We set α = 0.001 and s0 = 0 by default. We also set ηi = 1/(Ni(s, a) + 1), ϵ = 0.1/( Ni(s) + 1) and Q(0)(s, a) = 1, where Ni(s, a) and Ni(s) denote the number of visiting (s, a) pair and state s up to i-th iteration.