Optimal Pricing for the Competitive and Evolutionary Cloud Market

Authors: Bolei Xu, Tao Qin, Guoping Qiu, Tie-Yan Liu

IJCAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Numerical simulations demonstrate the effectiveness of our proposed approach. We carry out a large set of numerical experiments and show that our proposed approach can help a provider to achieve much better profit against his/her reactive opponents than other simple pricing policies.
Researcher Affiliation Collaboration Bolei Xu The University of Nottingham Ningbo China Bolei.Xu@nottingham.edu.cn Tao Qin Microsoft Research taoqin@microsoft.com Guoping Qiu The University of Nottingham Ningbo China Guoping.Qiu@nottingham.edu.cn Tie-Yan Liu Microsoft Research tyliu@microsoft.com
Pseudocode Yes Algorithm 1 Q-learning for a stationary market; Algorithm 2 Backward induction for an evolutionary market
Open Source Code No The paper does not provide any concrete access information (e.g., a URL or an explicit statement of code release) for its source code.
Open Datasets No The paper describes numerical simulations with specific parameter settings but does not refer to a publicly available dataset or provide access information for any dataset used in training or evaluation.
Dataset Splits No The paper describes numerical simulations and parameter settings but does not mention specific training, validation, or test dataset splits, as it does not use pre-existing datasets in the traditional sense.
Hardware Specification No The paper does not provide any specific hardware details (e.g., GPU/CPU models, memory specifications, or cloud instance types) used for running its numerical simulations.
Software Dependencies No The paper does not mention any specific software dependencies with version numbers that would be required to replicate the experiments (e.g., specific programming languages, libraries, or solvers with their versions).
Experiment Setup Yes Parameter Settings For the simulations in this subsection, we consider a market with K = 3 providers. ... We set the threshold price ω = 0.1 in all the scenarios. ... For the simulations, we set the initial population number N0 = 100, the saturated population number is N = 10, 000, and the evolution rate κ = 0.07. ... We assume the users marginal value θj derived from cloud service follows a uniform distribution supported on [1, 5] and the users demand follows an exponential distribution and set the parameter Λ in Eq. (2) to 2. We set β = 0.01 and η = 0.02 for provider s marginal cost function in Eq. (4). For the Q-learning algorithm, we set learning rate α = 0.1 1+0.1j which is decreasing with respect to the iteration step j and the stop criteria δ = 1.