Consider a discrete source producing a sequence of message letters from a finite alphabet. A single-letter distortion measure is given by a non-negative matrix (d ij ). The entryd ij measures the ?cost? or ?distortion? if letteriis reproduced at the receiver as letterj. The average distortion of a communications system (source-coder-noisy channel-decoder) is taken to bed= ? i.j P ij d ij whereP ij is the probability ofibeing reproduced asj. It is shown that there is a functionR(d) that measures the ?equivalent rate? of the source for a given level of distortion. For coding purposes where a leveldof distortion can be tolerated, the source acts like one with information rateR(d). Methods are given for calculatingR(d), and various properties discussed. Finally, generalizations to ergodic sources, to continuous sources, and to distortion measures involving blocks of letters are developed. In this paper a study is made of the problem of coding a discrete source of information, given afidelity criterionor ameasure of the distortionof the final recovered message at the receiving point relative to the actual transmitted message. In a particular case there might be a certain tolerable level of distortion as determined by this measure. It is desired to so encode the information that the maximum possible signaling rate is obtained without exceeding the tolerable distortion level. This work is an expansion and detailed elaboration of ideas presented earlier [1], with particular reference to the discrete case. We shall show that for a wide class of distortion measures and discrete sources of information there exists a functionR(d) (depending on the particular distortion measure and source) which measures, in a sense, the equivalent rateRof the source (in bits per letter produced) whendis the allowed distortion level. Methods will be given for evaluatingR(d) explicitly in certain simple cases and for evaluatingR(d) by a limiting process in more complex cases. The basic results are roughly that it is impossible to signal at a rate faster thanC/R(d) (source letters per second) over a memoryless channel of capacityC(bits per second) with a distortion measure less than or equal tod. On the other hand, by sufficiently long block codes it is possible to approach as closely as desired the rateC/R(d) with distortion leveld. Finally, some particular examples, using error probability per letter of message and other simple distortion measures, are worked out in detail.

Coding Theorems for a Discrete Source With a Fidelity CriterionInstitute of Radio Engineers, International Convention Record, vol. 7, 1959.

We provide an introduction to theoretical and practical aspects of Boosting and Ensemble learning, providing a useful reference for researchers in the field of Boosting as well as for those seeking to enter this fascinating area of research. We begin with a short background concerning the necessary learning theoretical foundations of weak learners and their linear combinations. We then point out the useful connection between Boosting and the Theory of Optimization, which facilitates the understanding of Boosting and later on enables us to move on to new Boosting algorithms, applicable to a broad spectrum of problems. In order to increase the relevance of the paper to practitioners, we have added remarks, pseudo code, "tricks of the trade", and algorithmic considerations where appropriate. Finally, we illustrate the usefulness of Boosting algorithms by giving an overview of some existing applications. The main ideas are illustrated on the problem of binary classification, although several extensions are discussed.

/pdf/an-introduction-to-boosting-and-leveraging-36mfqregqi.pdf

An introduction to boosting and leveraging

The goal of transfer learning is to improve the learning of a new target concept given knowledge of related source concept(s). We introduce the first boosting-based algorithms for transfer learning that apply to regression tasks. First, we describe two existing classification transfer algorithms, ExpBoost and TrAdaBoost, and show how they can be modified for regression. We then introduce extensions of these algorithms that improve performance significantly on controlled experiments in a wide range of test domains.

/pdf/boosting-for-regression-transfer-3707tc3dbk.pdf

Boosting for Regression Transfer

A market-based scheduling mechanism allocates resources indexed by time to alternative uses based on the bids of participating agents. Agents are typically interested in multiple time slots of the schedulable resource, with value determined by the earliest deadline by which they can complete their corresponding tasks. Despite the strong complementarities among slots induced by such preferences, it is often infeasible to deploy a mechanism that coordinates allocation across all time slots. We explore the case of separate, simultaneous markets for individual time slots, and the strategic problem it poses for bidding agents. Investigation of the straightforward bidding policy and its variants indicates that the efficacy of particular strategies depends critically on preferences and strategies of other agents, and that the strategy space is far too complex to yield to general game-theoretic analysis. For particular environments, however, it is often possible to derive constrained equilibria through evolutionary search methods.

/pdf/exploring-bidding-strategies-for-market-based-scheduling-4tt6l9op44.pdf

Exploring bidding strategies for market-based scheduling

Pay per click (PPC) auctions are used to sell positions in search engines. These auctions have gained increasing commercial importance, and many companies offer software to bid on these auctions. We present a trading agent for PPC auctions that is the first in our knowledge to use an explicit profit objective function. The agent creates a look-ahead plan of its desired bids, which allows it to exhibit intelligent behaviour including the ability to hold back money during expensive periods. We tested the agent in the latter part of 2003 in a live Overture auction. The agent generated four times the number of visits as human managers, in addition to reducing cost and variability.

/pdf/optimal-bidding-on-keyword-auctions-nc8o46nrdp.pdf

Optimal Bidding on Keyword Auctions

Auctions are becoming an increasingly popular method for transacting business, especially over the Internet. This article presents a general approach to building autonomous bidding agents to bid in multiple simultaneous auctions for interacting goods. A core component of our approach learns a model of the empirical price dynamics based on past data and uses the model to analytically calculate, to the greatest extent possible, optimal bids. We introduce a new and general boosting-based algorithm for conditional density estimation problems of this kind, i.e., supervised learning problems in which the goal is to estimate the entire conditional distribution of the real-valued label. This approach is fully implemented as ATTac-2001, a top-scoring agent in the second Trading Agent Competition (TAC-01). We present experiments demonstrating the effectiveness of our boosting-based price predictor relative to several reasonable alternatives.

/pdf/decision-theoretic-bidding-based-on-learned-density-models-3xkh7j4zi7.pdf

Decision-theoretic bidding based on learned density models in simultaneous, interacting auctions

In complicated, interacting auctions, a fundamental problem is the prediction of prices of goods in the auctions, and more broadly, the modeling of uncertainty regarding these prices. In this paper, we present a machine-learning approach to this problem. The technique is based on a new and general boosting-based algorithm for conditional density estimation problems of this kind, i.e., supervised learning problems in which the goal is to estimate the entire conditional distribution of the real-valued label. This algorithm, which we present in detail, is at the heart of , a top-scoring agent in the recent Trading Agent Competition (TAC-01). We describe how works, the results of the competition, and controlled experiments evaluating the effectiveness of price prediction in auctions.

/pdf/modeling-auction-price-uncertainty-using-boosting-based-1if4eg2vu0.pdf

Modeling Auction Price Uncertainty Using Boosting-based Conditional Density Estimation

Auctions are becoming an increasingly popular method for transacting business, especially over the Internet. This paper presents a general approach to building autonomous bidding agents to bid in multiple simultaneous auctions for interacting goods. The core of our approach is learning a model of the empirical price dynamics based on past data and using the model to analytically calculate, to the greatest extent possible, optimal bids. This approach is fully implemented as ATTac-2001, a top-scoring agent in the second Trading Agent Competition (TAC-01). ATTac-2001 uses boosting techniques to learn conditional distributions of auction clearing prices. We present experiments demonstrating the effectiveness of this predictor relative to several reasonable alternatives.

/pdf/attac-2001-a-learning-autonomous-bidding-agent-ti9d2v227w.pdf

ATTac-2001: A Learning, Autonomous Bidding Agent

Private computation using a PEZ dispenser

We introduce FAucS, a software testbed for studying automated agent bidding strategies in simulated auctions, specifically the United States FCC wireless frequency spectrum auctions. In addition to the complexity of these auctions, which provides ample opportunities for intelligent approaches to bidding, this type of auction has huge commercial importance, each bringing in billions of dollars to governments around the world. We implement straightforward sample agents in FAucS and use them to replicate known beneficial bidding strategies in this type of auction. We then discuss potential in-depth studies of autonomous bidding agent behaviors using FAucS. The main contribution of this work is the implementation, description, and empirical validation of the FAucS testbed. We present it as a challenging and promising AI research domain.

/pdf/faucs-an-fcc-spectrum-auction-simulator-for-autonomous-2zac9ccfrx.pdf

János A. Csirik

Papers

Decision-theoretic bidding based on learned density models in simultaneous, interacting auctions

Modeling Auction Price Uncertainty Using Boosting-based Conditional Density Estimation

ATTac-2001: A Learning, Autonomous Bidding Agent

Private computation using a PEZ dispenser

FAucS: An FCC Spectrum Auction Simulator for Autonomous Bidding Agents