The distance metric plays an important role in nearest neighbor (NN) classification. Usually the Euclidean distance metric is assumed or a Mahalanobis distance metric is optimized to improve the NN performance. In this paper, we study the problem of embedding arbitrary metric spaces into a Euclidean space with the goal to improve the accuracy of the NN classifier. We propose a solution by appealing to the framework of regularization in a reproducing kernel Hilbert space and prove a representer-like theorem for NN classification. The embedding function is then determined by solving a semidefinite program which has an interesting connection to the soft-margin linear binary support vector machine classifier. Although the main focus of this paper is to present a general, theoretical framework for metric embedding in a NN setting, we demonstrate the performance of the proposed method on some benchmark datasets and show that it performs better than the Mahalanobis metric learning algorithm in terms of leave-one-out and generalization errors.

We consider the problem of computing a lightest derivation of a global structure using a set of weighted rules. A large variety of inference problems in AI can be formulated in this framework. We generalize A* search and heuristics derived from abstractions to a broad class of lightest derivation problems. We also describe a new algorithm that searches for lightest derivations using a hierarchy of abstractions. Our generalization of A* gives a new algorithm for searching AND/OR graphs in a bottom-up fashion. We discuss how the algorithms described here provide a general architecture for addressing the pipeline problem --- the problem of passing information back and forth between various stages of processing in a perceptual system. We consider examples in computer vision and natural language processing. We apply the hierarchical search algorithm to the problem of estimating the boundaries of convex objects in grayscale images and compare it to other search methods. A second set of experiments demonstrate the use of a new compositional model for finding salient curves in images.

Conformal prediction uses past experience to determine precise levels of confidence in new predictions. Given an error probability $\epsilon$, together with a method that makes a prediction $\hat{y}$ of a label $y$, it produces a set of labels, typically containing $\hat{y}$, that also contains $y$ with probability $1-\epsilon$. Conformal prediction can be applied to any method for producing $\hat{y}$: a nearest-neighbor method, a support-vector machine, ridge regression, etc. Conformal prediction is designed for an on-line setting in which labels are predicted successively, each one being revealed before the next is predicted. The most novel and valuable feature of conformal prediction is that if the successive examples are sampled independently from the same distribution, then the successive predictions will be right $1-\epsilon$ of the time, even though they are based on an accumulating dataset rather than on independent datasets. In addition to the model under which successive examples are sampled independently, other on-line compression models can also use conformal prediction. The widely used Gaussian linear model is one of these. This tutorial presents a self-contained account of the theory of conformal prediction and works through several numerical examples. A more comprehensive treatment of the topic is provided in "Algorithmic Learning in a Random World", by Vladimir Vovk, Alex Gammerman, and Glenn Shafer (Springer, 2005).

By early July, each paper had been assigned to one supervisor SPC member and one PC member. The algorithm recorded the justifications for each assignment in terms of the specific bid and keyword match. When completed, the reviews were and Its Benefits to Society." The tutorial program was Hyderabad, India, January 6-12, 2007. At the chaired by Cynthia Braezeal. More The theme of the conference was "AI Figure 2 shows the distribution of their course work.

'We should be worried, for aaai.org/aitopics/ We are Please note that: (1) an excerpt may not my understanding, I visited USC's'We need to tell the the fact that an item has been selected does University of Massachusetts in Amherst, Robot Wars -- An Attempt to Build an with many of the programs in the omy -- that prompted interest in the technology, Ethical Robotic Soldier. 'We are Technology, in Atlanta, is developing a set those tested uses the sort of artificial intelligence studying the application of the RAHS concepts of rules of engagement for battlefield technology that encourages highlevel and tools to the social, and economic robots to ensure that their use of lethal interactivity.... Call me an industry and financial domains,' Nathan wrote force follows the rules of ethics. In other cheerleader, but what I see at [William in an email interview." Kim conscience.... His approach is to create suggests that computers are already helping Yoon-mi. April 28, 2007 what he calls a'multidimensional mathematical students learn and will become increasingly (www.koreaherald.co.kr). "To literally live decision space of possible behavior important year by year." with robots, that are highly likely to become actions'.... Arkin has started to survey policy Search Engine Spawned from Antiterrorism more intelligent and physically closer makers, the public, researchers and military Efforts Finds Place in Business. "Artificial-intelligence-based that will prevent robots from doing harm Computer Science Takes Steps to Bring search technology to people, and block humans from taking Women to the Fold.

Summer calendar of AI events.

Approximately November 9-11, at the Westin Arlington and IAAI-08, please see www.aaai.org/

The Dialogue on Dialogues workshop was organized as a satellite event at the Interspeech 2006 conference in Pittsburgh, Pennsylvania, and it was held on September 17, 2006, immediately before the main conference. It was planned and coordinated by Michael McTear (University of Ulster, UK), Kristiina Jokinen (University of Helsinki, Finland), and James A. Larson (Portland State University, USA). The one-day workshop involved more than 40 participants from Europe, the United States, Australia, and Japan.

Will we see autonomous humanoid robots that play (and win) soccer against the human soccer world champion in the year 2050? This question is not easy to answer, and the idea is quite visionary. However, this is the goal of the RoboCup Federation. There are serious research questions that have to be tackled behind the scenes of a soccer game: perception, decision making, action selection, hardware design, materials, energy, and more. RoboCup is also about the nature of intelligence, and playing soccer acts as a performance measure of systems that contain artificial intelligence -- in much the same way chess has been used over the last century. This article outlines the current situation following 10 years of research with reference to the results of the 2006 World Championship in Bremen, Germany, and discusses future challenges.

Schultz will speak on moving toward Intelligence Conference (IAAI-Vancouver. Oren Etzioni (University of Washington of New South Wales), distinguished for and Farecast) is known for his his work in automated reasoning, will technical work in intelligent agents, speak on representing and reasoning data mining and web search. Coolest Web Sites" and PC World's "20 progress report on logic for automated Applications of AI Conference Intelligent Arts) was the technical director and agents is another research area that experienced designer for Electronic Art's FIFA franchise He is obsessed Careful thought was put into inviting College Park) is a prominent with creating the illusion of life world class speakers to this year's conference. Getoor will speak on graph application of AI. In his talk, entitled Alan Mackworth (University of British identification, specifically methods "Big'A,' Small'I': Smart Ends from Columbia). For the technical AAAI-07 that transform an observed input Simple Means," Brown will cover various conference, speakers include Alan graph into an inferred output graph.