We study online learning in Boolean domains using kernels which capture feature expansions equivalent to using conjunctions over basic features. We demonstrate a tradeoff between the computational efficiency with which these kernels can be computed and the generalization ability of the resulting classifier. We first describe several kernel functions which capture either limited forms of conjunctions or all conjunctions. We show that these kernels can be used to efficiently run the Perceptron algorithm over an exponential number of conjunctions; however we also prove that using such kernels the Perceptron algorithm can make an exponential number of mistakes even when learning simple functions. We also consider an analogous use of kernel functions to run the multiplicative-update Winnow algorithm over an expanded feature space of exponentially many conjunctions. While known upper bounds imply that Winnow can learn DNF formulae with a polynomial mistake bound in this setting, we prove that it is computationally hard to simulate Winnow's behavior for learning DNF over such a feature set, and thus that such kernel functions for Winnow are not efficiently computable.

In this paper we show that online algorithms for classification and regression can be naturally used to obtain hypotheses with good datadependent tail bounds on their risk. Our results are proven without requiring complicated concentration-of-measure arguments and they hold for arbitrary online learning algorithms. Furthermore, when applied to concrete online algorithms, our results yield tail bounds that in many cases are comparable or better than the best known bounds.

We study online learning in Boolean domains using kernels which capture feature expansions equivalent to using conjunctions over basic features. We demonstrate a tradeoff between the computational efficiency with which these kernels can be computed and the generalization ability of the resulting classifier. We first describe several kernel functions which capture either limited forms of conjunctions or all conjunctions. We show that these kernels can be used to efficiently run the Perceptron algorithm over an exponential number of conjunctions; however we also prove that using such kernels the Perceptron algorithm can make an exponential number of mistakes even when learning simple functions. We also consider an analogous use of kernel functions to run the multiplicative-update Winnow algorithm over an expanded feature space of exponentially many conjunctions. While known upper bounds imply that Winnow can learn DNF formulae with a polynomial mistake bound in this setting, we prove that it is computationally hard to simulate Winnow's behavior for learning DNF over such a feature set, and thus that such kernel functions for Winnow are not efficiently computable.

We study online learning in Boolean domains using kernels which capture featureexpansions equivalent to using conjunctions over basic features. Wedemonstrate a tradeoff between the computational efficiency with which these kernels can be computed and the generalization ability ofthe resulting classifier. We first describe several kernel functions which capture either limited forms of conjunctions or all conjunctions. We show that these kernels can be used to efficiently run the Perceptron algorithmover an exponential number of conjunctions; however we also prove that using such kernels the Perceptron algorithm can make an exponential number of mistakes even when learning simple functions. Wealso consider an analogous use of kernel functions to run the multiplicative-update Winnow algorithm over an expanded feature space of exponentially many conjunctions. While known upper bounds imply that Winnow can learn DNF formulae with a polynomial mistake bound in this setting, we prove that it is computationally hard to simulate Winnow's behaviorfor learning DNF over such a feature set, and thus that such kernel functions for Winnow are not efficiently computable.

In this paper we show that online algorithms for classification and regression canbe naturally used to obtain hypotheses with good datadependent tailbounds on their risk. Our results are proven without requiring complicated concentration-of-measure arguments and they hold for arbitrary online learning algorithms. Furthermore, when applied to concrete online algorithms, our results yield tail bounds that in many cases are comparable or better than the best known bounds.

Originally introduced them together in a synergistic manner has resulted by Clausewitz in his classical work On in faster progress for each of them. War (1976), the center of gravity is now understood Moreover, it offers a new perspective on how to as representing "those characteristics, capabilities, combine research in AI with research in a specialized or localities from which a military domain and with the development force derives its freedom of action, physical and deployment of prototype systems in education strength, or will to fight" (Joint Chiefs of Staff and practice.

SOAR is one of the oldest and largest AI development efforts, starting formally in 1983. It has also been proposed as a unified theory of cognition (Newell 1990). Most of its current development is as an AI programming language, which was evident at the Twenty-Second SOAR Workshop held at Soar Technology near the University of Michigan in Ann Arbor on 1-2 June 2002.

The RoboCup Rescue Physical Agent League Competition was held in the summer of 2001 in conjunction with the AAAI Mobile Robot Competition Urban Search and Rescue event, eerily preceding the September 11 World Trade Center (WTC) disaster. Four teams responded to the WTC disaster through the auspices of the Center for Robot-Assisted Search and Rescue (CRASAR), directed by John Blitch. The four teams were Foster- Miller and iRobot (both robot manufacturers from the Boston area), the United States Navy's Space Warfare Center (SPAWAR) group from San Diego, and the University of South Florida (USF). Blitch, through his position as program manager for the Defense Advanced Research Projects Agency (DARPA) Tactical Mobile Robots Program, was a supporter of the competition; he also served as a member of the rules committee and a judge. USF participated by chairing the rules committee, judging, assisting with the logistics, providing commentary, and demonstrating tethered and wireless robots whenever entrants had to skip around during the competition. Based on our experiences and history, we were asked to comment on the validity of the competition. The CRASAR collective experience suggests that most of the basic rules of the competition matched reality because the rules accurately reflected deployment scenarios, but the National Institute of Standards and Technology (NIST) Standard Test Course, and hardware or software approaches forwarded by competitors in last summer's event, missed the mark. This article briefly reviews the types of robots and missions used by CRASAR at the WTC site, then discusses the robotassisted search and rescue effort in terms of lessons for the competition.

They express both thoughts and California (USC)/Information Sciences Institute emotions; emotional expression is important to (ISI) is to develop new technologies that portray characteristics of enthusiasm and empathy promote effective learning and increase learner that are important for human teachers. These technologies are intended They are knowledgeable about the subject matter to result in interactive learning materials that being learned, of pedagogical strategies, and support the learning process and that complement also have knowledge about how to find and and enhance existing technologies relevant obtain relevant knowledge from available to learning such as the World Wide Web. Our work draws significant inspiration from Figure 1 shows one of the guidebots that we human learning and teaching. We piece of equipment called a high-pressure air seek a better understanding of the characteristics compressor aboard United States Navy ships. As learners view instructional materials, guidebots can provide useful commentary on these materials.

BDDplan was created to perform certain reasoning processes in the fluent calculus, a flexible framework for reasoning about action and change based on first-order logic with equality (plus some second-order extensions in some cases). The reasoning is done by mapping the problems into propositional logic, which, in turn, can be implemented as operations on binary decision diagrams (BDDs).