Biased Random-key Genetic Algorithms For The Winner Determination Problem In Combinatorial Auctions.

Abstract In this paper, we address the problem of picking a subset of bids in a general combinatorial auction so as to maximize the overall profit using the first-price model. This winner determination problem assumes that a single bidding round is held to determine both the winners and prices to be paid. We introduce six variants of biased random-key genetic algorithms for this problem. Three of them use a novel initialization technique that makes use of solutions of intermediate linear programming relaxations of an exact mixed integer-linear programming model as initial chromosomes of the population. An experimental evaluation compares the effectiveness of the proposed algorithms with the standard mixed linear integer programming formulation, a specialized exact algorithm, and the best-performing heuristics proposed for this problem. The proposed algorithms are competitive and offer strong results, mainly for large-scale auctions.

Egalitarian behaviour in multi unit combinatorial auctions

En entornos donde los recursos son precederos y la asignación de recursos se repite en el tiempo con el mismo conjunto o un conjunto muy similar de agentes, las subastas recurrentes pueden ser utilizadas. Una subasta recurrente es una secuencia de subastas donde el resultado de una subasta puede influenciar en las siguientes. De todas formas, este tipo de subastas tienen problemas particulares cuando la riqueza de los agentes esta desequilibrada y los recursos son precederos. En esta tesis se proponen algunos mecanismos justos o equitativos para minimizar los efectos de estos problemas. En una subasta recurrente una solución justa significa que todos los participantes consiguen a largo plazo sus objetivos en el mismo grado o en el grado más parecido posible, independientemente de su riqueza. Hemos demostrado experimentalmente que la inclusión de justicia incentiva a los bidders en permanecer en la subasta minimizando los problemas de las subastas recurrentes.

GRASP with Hybrid Path Relinking for Bi-Objective Winner Determination in Combinatorial Transportation Auctions

The procurement of transportation services via large-scale combinatorial auctions involves a couple of complex decisions whose outcome highly influences the performance of the tender process. This paper examines the shipper's task of selecting a subset of the submitted bids which efficiently trades off total procurement cost against expected carrier performance. To solve this bi-objective winner determination problem, we propose a Pareto-based greedy randomized adaptive search procedure (GRASP). As a post-optimizer we use a path relinking procedure which is hybridized with branch-and-bound. Several variants of this algorithm are evaluated by means of artificial test instances which comply with important real-world characteristics. The two best variants prove superior to a previously published Pareto-based evolutionary algorithm.

Leilões combinados

Mestrado em Controlo de Gestão e dos Negócios

Representando o vendedor e a sociedade : mecanismos com dois critérios para leilões combinatórios com demandas unitárias

This work focuses on obtaining truthful mechanisms that aim at maximizing both the revenue and the economic efficiency (social welfare) for the unitdemand combinatorial auction problem (UDCAP), in which a set of k items is auctioned to a set of n consumers. Although each consumer bids on all items, no consumer can purchase more than one item in the UDCAP. We present a framework for devising poly-time randomized competitive truthful mechanisms that can be used to either favor economic efficiency or revenue.

Effiziente, operative Torbelegung mittels kombinatorischer Auktionen

Das operative Torbelegungsproblem (TBP) z. B. an einem Distributions- oder Cross-dockingzentrum ist ein logistisches Problem, bei dem es gilt, an- und abfahrende Fahrzeuge zeitlich und räumlich so auf die Warenein- und -ausgangstore zu verteilen, dass eine mög-lichst kostengünstige Abfertigung ermöglicht wird. Bisherige Arbeiten, die sich mit dem TBP beschäftigen, lassen Aspekte der Kooperation außer Acht. Dieser Beitrag stellt ein Verfahren vor, durch das der Nachteil einseitig optimaler Torbelegungen überwunden werden kann. Dabei wird auf das Mittel der kombinatorischen Auktionen zurückgegriffen und das TBP als Allokationsproblem modelliert, bei dem Frachtführer um Bündel konsekutiver Einheitszeit-intervalle an den Toren konkurrieren. Mittels eines Vickrey-Clarke-Groves-Mechanismus wird einerseits die Anreizkompatibilität, andererseits die individuelle Rationalität des Auk-tionsverfahrens sichergestellt. Das Verfahren wurde in ILOG OPL Studio 3.6.1 implemen-tiert und die durch Testdaten gewonnenen Ergebnisse zeigen, dass die Laufzeiten gering genug sind, um das Verfahren für die operative (kurzfristige) Planung einzusetzen und so transportlogistische Prozesse für alle Beteiligten wirtschaftlicher zu gestalten.

Social interaction as a heuristic for combinatorial optimization problems

We investigate the performance of a variant of Axelrod's model for dissemination of culture-the Adaptive Culture Heuristic (ACH)-on solving an NP-Complete optimization problem, namely, the classification of binary input patterns of size F by a Boolean Binary Perceptron. In this heuristic, N agents, characterized by binary strings of length F which represent possible solutions to the optimization problem, are fixed at the sites of a square lattice and interact with their nearest neighbors only. The interactions are such that the agents' strings (or cultures) become more similar to the low-cost strings of their neighbors resulting in the dissemination of these strings across the lattice. Eventually the dynamics freezes into a homogeneous absorbing configuration in which all agents exhibit identical solutions to the optimization problem. We find through extensive simulations that the probability of finding the optimal solution is a function of the reduced variable F/N(1/4) so that the number of agents must increase with the fourth power of the problem size, N proportional to F(4), to guarantee a fixed probability of success. In this case, we find that the relaxation time to reach an absorbing configuration scales with F(6) which can be interpreted as the overall computational cost of the ACH to find an optimal set of weights for a Boolean binary perceptron, given a fixed probability of success.

Simultaneous pooled auctions

Suppose a seller wants to sell k similar or identical objects and there are n > k potential buyers. Suppose that each buyer wants only one object. In this case, we suggest the use of a simultaneous auction that would work as follows. Players are asked to submit sealed bids for one object. The individual with the highest bid chooses an object first; the individual with the second-highest bid chooses the next object; and this process continues until the individual with the kth highest bid receives the last object. Each individual pays the equivalent to his or her bid. When objects are identical, we show that the proposed auction generates the same revenue as a first-price sealed-bid sequential auction. When objects are perfectly correlated, there is no known solution for sequential auctions, whereas we can characterize bidding strategies in the proposed auction. Moreover, the proposed auction is optimal (given an appropriately chosen reserve price), and it may be easier and cheaper to run than a sequential auction.

A simple analysis of the US emission permits auctions

We examine a stylized version of EPA auctions when agents know the list of values of sellers and buyers. There are inefficient equilibria where no goods are traded and efficient equilibria where all exchange occurs at a uniform price. We also provide examples under incomplete information when the uniform price equilibrium holds and when it does not hold. (C) 1999 Elsevier Science S.A. All rights reserved. JEL classification: D44; Q29.

Encoding combinatorial libraries: A novel application of fluorescent silica colloids

A major challenge associated with using large chemical libraries synthesized on microscopic solid support beads is the rapid discrimination of individual compounds in these libraries. This challenge can be overcome by encoding the beads with 1 mum silica colloidal particles (reporters) that contain specific and identifiable combinations of fluorescent byes. The colored bar code generated on support beads during combinatorial library synthesis can be easily, rapidly, and inexpensively decoded through the use of fluorescence microscopy. All reporters are precoated with polyelectrolytes [poly(acrylic acid), PAA, poly(sodium 4-styrenesulfonate PSSS, polyethylenimine, PEI, and/or poly(diallyldimethylammonium chloride), PDADMAC] with the aim of enhancing surface charge, promoting electrostatic attraction to the bead, and facilitating polymer bridging between the bead and reporter for permanent adhesion. As shown in this article, reporters coated with polyelectrolytes clearly outperform uncoated reporters with regard to quantity of attached reporters per bead (54 +/- 23 in 2500 mum(2) area for PEI/PAA coated and 11 +/- 6 for uncoated reporters) and minimization of cross-contamination (1 red reporter in 2500 mum(2) area of green-labeled bead for PEI/PAA coated and 26 +/- 15 red reporters on green-labeled beads for uncoated reporters after 10 days). Examination of various polyelectrolyte systems shows that the magnitude of the xi -potential of polyelectrolyte-coated reporters (-64 mV for PDADMAC/PSSS and -42 mV for PEI/PAA-coated reporters) has no correlation with the number of reporters that adhere to the solid support beads (21 +/- 16 in 2500 mum(2) area for PDADMAC/PSSS and 54 +/- 23 for PEI/PAA-coated reporters). The contribution of polymer bridging to the adhesion has a far greater influence than electrostatic attraction and is demonstrated by modification of the polyelectrolyte multilayers using gamma irradiation of precoated reporters either in aqueous solution or in polyelectrolyte solution.

Hybrid auctions

In this paper we analyze a hybrid auction that combines a first-price and a Vickrey auction. We show that this auction may generate more expected revenue than a standard first-price auction. (C) 2002 Elsevier Science B.V. All rights reserved.

Auction price anomalies: Evidence from wool auctions in Australia

Using detailed Australian wool auction data we test for further evidence of pricing anomalies at sequential auctions. We find that an anomaly frequently exists and order is frequently endogenously determined. Moreover, prices increase through some sales and decrease through others. We examine whether it is possible to explain the variation in the anomaly across sales and conclude that there is no systematic relationship between the direction of the price anomaly and the characteristics of the wool or the auction. We do, however, find evidence that an anomaly, is more likely in longer sales.

Auctions with synergies and asymmetric buyers

In this paper we consider sequential auctions with synergies where one player wants two objects and the remaining players want one object each. We show that expected prices may not necessarily decrease as predicted by Branco [Econ. Lett. 54 (1997) 159]. Indeed we show that expected prices can actually increase. (C) 2004 Elsevier B.V All rights reserved.

Reserve price commitments in auctions

It is well known that the optimal auction-one that maximizes the seller's expected revenue-can be implemented using a standard auction format with a suitably chosen reserve price. This reserve price is above the seller's value of retaining the object and the mechanism requires a commitment not to sell the object below the reserve. This commitment is what makes the reserve valuable to the seller. However, in practice, a reserve price commits the seller to sell the object if the reserve is reached, but does not commit her to withhold the object from sale if bidding falls short of the reserve. In this note we investigate whether reserve prices remain valuable for the seller when she may negotiate with the highest bidder if the reserve is not met. We show that the value of the reserve price may be completely undermined if the seller is a sufficiently weak bargainer. (c) 2004 Elsevier B.V. All rights reserved.