Autonomous Bidding Agents: Strategies and Lessons from the by Michael P. Wellman

By Michael P. Wellman

E-commerce more and more presents possibilities for self sustaining bidding brokers: desktop courses that bid in digital markets with out direct human intervention. automatic bidding ideas for an public sale of a unmarried strong with a identified valuation are really ordinary; designing ideas for simultaneous auctions with interdependent valuations is a extra complicated venture. This publication offers algorithmic advances and approach rules inside an built-in bidding agent structure that experience emerged from fresh paintings during this fast-growing region of study in academia and undefined. The authors learn a number of novel bidding methods that built from the buying and selling Agent pageant (TAC), held every year for the reason that 2000. The benchmark problem for competing agents--to purchase and promote a number of items with interdependent valuations in simultaneous auctions of other types--encourages rivals to use leading edge ideas to a typical activity. The publication strains the evolution of TAC and follows chosen brokers from perception via numerous competitions, proposing and studying specific algorithms built for independent bidding. self reliant Bidding brokers offers the 1st built-in remedy of equipment during this swiftly constructing area of AI. The authors--who brought TAC and created a few of its so much winning agents--offer either an outline of present examine and new effects. Michael P. Wellman is Professor of machine technological know-how and Engineering and member of the factitious Intelligence Laboratory on the college of Michigan, Ann Arbor. Amy Greenwald is Assistant Professor of desktop technology at Brown college. Peter Stone is Assistant Professor of laptop Sciences, Alfred P. Sloan learn Fellow, and Director of the training brokers team on the college of Texas, Austin. he's the recipient of the overseas Joint convention on man made Intelligence (IJCAI) 2007 desktops and idea Award.

Show description

Read or Download Autonomous Bidding Agents: Strategies and Lessons from the Trading Agent Competition PDF

Similar intelligence & semantics books

Evolutionary Computation in Practice

This ebook is loaded with examples within which laptop scientists and engineers have used evolutionary computation - courses that mimic usual evolution - to resolve actual difficulties. They aren t summary, mathematically in depth papers, yet bills of fixing vital difficulties, together with information from the authors on easy methods to keep away from universal pitfalls, maximize the effectiveness and potency of the quest approach, and plenty of different sensible feedback.

Feedforward Neural Network Methodology (Springer Series in Statistics)

This decade has obvious an explosive progress in computational velocity and reminiscence and a fast enrichment in our figuring out of man-made neural networks. those components offer structures engineers and statisticians having the ability to construct types of actual, monetary, and information-based time sequence and signs.

Artificial Intelligence for Humans, Volume 2: Nature-Inspired Algorithms

Nature could be a nice resource of idea for man made intelligence algorithms simply because its expertise is significantly extra complicated than our personal. between its wonders are robust AI, nanotechnology, and complex robotics. Nature can for that reason function a advisor for real-life challenge fixing. during this publication, you are going to come upon algorithms inspired through ants, bees, genomes, birds, and cells that offer sensible equipment for lots of kinds of AI occasions.

Extra resources for Autonomous Bidding Agents: Strategies and Lessons from the Trading Agent Competition

Example text

Ng }. Finally, we define pg to be qg permuted according to σ. 4 We call the output of this procedure unified pricelines, the set of which we denote by P . Note that these operations can be carried out in polynomial time (see Algorithm 2). Specifically, the time complexity of Unify(G, N, P, Π) (and of this reduction overall) is O(|G|K log K), where K = maxg Ng . Algorithm 2 Unify(G, N, P, Π) 1: for all g ∈ G do 2: for k = 1 . . , σ = {1 → 2, 2 → 1} sorts 20, 10 accordingly} 6: for k = 1 . . Ng do 7: let k = σg (k) 8: let pgk = qgk 9: end for 10: end for 11: return P For example, say there are four units of some good g˜ on the market with buyer priceline pg˜ = 0, 1, 10, 30 and seller priceline πg˜ = 40, 20, 0, 0 .

4 Visualization for Game 7321 of the TAC-01 finals. Bars represent bids, with height proportional to offer price and varying shades encoding the respective bidding agents. 24 Chapter 2 The absence of skyrocketing hotel prices cleared the way for other strategic issues to come to the fore in TAC-01. , 2003b] between the approaches taken by the two agents finishing at the top of the standings in the TAC-01 finals, ATTac and livingagents. 4) on all available goods. ATTac’s priceprediction module uses machine-learning techniques (see Chapter 6) to generate distributions over hotel closing prices.

On the other hand, if price projection involves complex computation or if bid determination involves difficult optimization, the agent may devote nearly the whole minute to these tasks, ignoring changes in other markets that occur during the computation. In general, agent designers trade off between the length of computation and the timeliness of information; or they construct flexible procedures that accommodate information updates within the computation process. 36 Chapter 3 Step 2: Model-Based Prediction of Prices and Holdings The update step defines the agent’s view of the current market state.

Download PDF sample

Rated 4.35 of 5 – based on 30 votes