A Decomposition of Forecast Error in Prediction Markets

Authors: Miro Dudik, Sebastien Lahaie, Ryan M. Rogers, Jennifer Wortman Vaughan

NeurIPS 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We demonstrate via numerical simulations that these bounds are tight. Our results yield new insights into the question of how to set the market s liquidity parameter and into the forecasting benefits of enforcing coherent prices across securities. (...) 7 Numerical Experiments We evaluate the tightness of our theoretical bounds via numerical simulation. We consider a complete market over K = 5 securities and simulate N = 10 traders with risk aversion coefficients equal to 1. (...) In Fig. 1 (center) we plot the bias µ (b; C) µ as a function of b for both LMSR and IND. We compare this with the theoretical approximation µ (b; C) µ b( a/N) HT ( µ) C ( µ) from Theorem 5.6.
Researcher Affiliation Collaboration Miroslav Dudík Microsoft Research, New York, NY mdudik@microsoft.com Sébastien Lahaie Google, New York, NY slahaie@google.com Ryan Rogers University of Pennsylvania, Philadelphia, PA rrogers386@gmail.com Jennifer Wortman Vaughan Microsoft Research, New York, NY jenn@microsoft.com
Pseudocode No The paper presents theoretical models, equations, and mathematical derivations, but it does not include any pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any statement about open-sourcing code for the described methodology, nor does it include links to a code repository.
Open Datasets No The paper states, 'We consider a complete market over K = 5 securities and simulate N = 10 traders with risk aversion coefficients equal to 1.' This indicates a simulation study rather than the use of a publicly available dataset. No public dataset source or access information is provided.
Dataset Splits No The paper describes a simulation setup rather than using an existing dataset with defined splits. There is no mention of training, validation, or test splits for data, or cross-validation.
Hardware Specification No The paper mentions that 'simulations are tractable' but does not provide any specific details about the hardware (e.g., CPU, GPU models, memory) used for running the experiments.
Software Dependencies No The paper describes the mathematical models and simulation setup but does not specify any software dependencies (e.g., programming languages, libraries, or solvers) with version numbers.
Experiment Setup Yes We consider a complete market over K = 5 securities and simulate N = 10 traders with risk aversion coefficients equal to 1. (...) We fix the ground-truth natural parameter θtrue and independently sample the belief θi of each trader from Normal(θtrue, σ2IK), with σ = 5. We consider a single-peaked ground truth distribution with θtrue 1 = log(1 ν(K 1)) and θtrue k = log ν for k = 1, with ν = 0.02. Trading is simulated according to the all-security dynamics (ASD) as described at the start of Section 6.