Differential Privacy for Stackelberg Games

Authors: Ferdinando Fioretto, Lesia Mitridati, Pascal Van Hentenryck

IJCAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on several gas and electricity market benchmarks based on a real case study demonstrate the effectiveness of the proposed approach. A full version of this paper [Fioretto et al., 2020b] contains complete proofs and additional discussion on the motivating application. (...) The theoretical guarantees ensure differential privacy and near optimality, while the experimental results validate the approach on a real test case for the coordination of electricity and natural gas markets in the Northeastern United States [Byeon and Van Hentenryck, 2019]. (...) 7 Experimental Evaluation The performance of the proposed PPSM is illustrated on the motivation problem introduced in Section 2.
Researcher Affiliation Academia Ferdinando Fioretto1 , Lesia Mitridati2 and Pascal Van Hentenryck2 1Syracuse University 2Georgia Institute of Technology
Pseudocode No The paper describes the PPSM mechanism using numbered steps ([1], [2a], etc.) in prose, but it does not provide formal pseudocode or an algorithm block.
Open Source Code No The paper does not provide any statement or link indicating that the source code for the described methodology is publicly available.
Open Datasets Yes The PPSM is evaluated on a test system representing the joint natural gas and electricity systems in the Northeastern US [Byeon and Van Hentenryck, 2019].
Dataset Splits No The paper evaluates on a 'test system' and generates 'repetitions for each test case', but it does not specify explicit train/validation/test dataset splits (e.g., percentages or counts) or reference standard predefined splits for reproducibility.
Hardware Specification No The paper does not provide specific details about the hardware used to run the experiments (e.g., CPU, GPU models, or memory specifications).
Software Dependencies No The paper mentions 'CPLEX' as a solver but does not provide a version number for it or for any other software dependencies. 'Using the equivalent Karush-Kuhn-Tucker (KKT) conditions of the linear lower-level problem (5d) and the Fortuny-Amat linearization, this bilevel problem can be recast as a mixed-integer second-order cone program (MISOCP) [Gabriel et al., 2012]. The resolution of the privacy-preserving demand profiles (phases [1] and [2] of PPSM) takes less than 30s for any of the instances.' (Implicitly, a solver for MISOCP like CPLEX might be used but no version is given).
Experiment Setup Yes The electricity demand profile is uniformly increased by a stress factor ranging from 30% to 60%, and the gas demand profile is increased by a stress factor ranging from 10% to 130%, producing increasingly stressed and difficult operating conditions. (...) The experiments compare the proposed PPSM to a version (PPSMp) that omits the fidelity constraint on the dual variables (5c). Both versions are compared with the standard Laplace mechanism for varying values of the privacy parameter α P t0.1, 1, 10u ˆ 102 MWh, and the fidelity parameters ηp ηd P t0.01, 0.1, 10.0u% of the original objective value of the GM p Og q and gas prices pyg q, respectively.