TacTex’13: A Champion Adaptive Power Trading Agent

Authors: Daniel Urieli, Peter Stone

AAAI 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental This paper describes the constituent components of TACTEX 13 and examines its success through analysis of competition results and subsequent controlled experiments.
Researcher Affiliation Academia Daniel Urieli and Peter Stone Dept. of Computer Science The University of Texas at Austin Austin, TX 78712 USA {urieli,pstone}@cs.utexas.edu
Pseudocode Yes Algorithm 1 Online RL Wholesale Market Strategy; Algorithm 2 Utility-based Tariff Market Strategy; Algorithm 3 Estimate Utility(tariff); Algorithm 4 Predict Change In Subscriptions(tariff)
Open Source Code No The paper states that Power TAC is an “open-source simulation platform,” but it does not provide any statement or link indicating that the source code for TACTEX 13 itself is open-source or publicly available.
Open Datasets No The paper mentions that “Power TAC uses real weather data and forecasts” and that games are “generated by loading a set of random-number seeds that initialize the random number generators of the simulation, and a weather data file.” However, it does not provide specific access information (link, citation for dataset) for a publicly available dataset used for TACTEX's training or evaluation.
Dataset Splits No The paper does not explicitly provide training/validation/test dataset splits for the evaluation of TACTEX 13. It describes competition scenarios (2-agent, 4-agent, 7-agent games) and controlled experiments (round-robin tournament), which serve as evaluation settings rather than formal dataset splits.
Hardware Specification No The paper does not specify any hardware used for running the experiments or for the Power TAC simulation platform.
Software Dependencies No The paper mentions Power TAC is an “open-source simulation platform” but does not provide specific version numbers for any software dependencies used in the implementation or experimentation of TACTEX 13.
Experiment Setup Yes TACTEX's wholesale market bidding strategy: 'The number 6 was chosen to trade off quick learning with reasonable density estimations.' TACTEX's tariff market strategy: 'The length of the horizon over which utility is estimated is one week (7 24 = 168 timeslots)'. 'generating a set of 100 fixed-rate candidate tariffs'. Controlled experiments: 'each given combination of agents was tested over a fixed set of 200 full games. Each game takes about 2 hours of real-time (about 60 days of simulated time), and was generated by loading a set of random-number seeds that initialize the random number generators of the simulation, and a weather data file'.