Ge, Ruifang. "Learning Semantic Parsers Using Statistical Syntactic Parsing Techniques." The University of Texas at Austin, Department of Computer Sciences. AI Technical Report 06-327. February 13, 2006. 41 pages.
Most recent work on semantic analysis of natural language has focused on "shallow'' semantics such as word-sense disambiguation and semantic role labeling. Our work addresses a more ambitious task we call semantic parsing where natural language sentences are mapped to complete formal meaning representations. We present our system Scissor based on a statistical parser that generates a semantically-augmented parse tree (SAPT), in which each internal node has both a syntactic and semantic label. A compositional-semantics procedure is then used to map the augmented parse tree into a final meaning representation. Training the system requires sentences annotated with augmented parse trees. We evaluate the system in two domains, a natural-language database interface and an interpreter for coaching instructions in robotic soccer. We present experimental results demonstrating that Scissor produces more accurate semantic representations than several previous approaches on long sentences.In the future, we intend to pursue several directions in developing more accurate semantic parsing algorithms and automating the annotation process. This work will involve exploring alternative tree representations for better generalization in parsing. We also plan to apply discriminative reranking methods to semantic parsing, which allows exploring arbitrary, potentially correlated features not usable by the baseline learner. We also propose to design a method for automating the SAPT-generation process to alleviate the extra annotation work currently required for training Scissor. Finally, we will investigate the impact of different statistical syntactic parsers on semantic parsing using the automated SAPT-generation process.
D'Silva, Thomas. "Applying the Policy Gradient Algorithm to Learn Stable Robot Locomotion." The University of Texas at Austin, Department of Computer Sciences. AI Technical Report 06-328. February 20, 2006. 6 pages.
In the RoboSoccer domain, a fast gait is an important component of a successful team. A robot with a fast gait often has body movements that cause unsteady camera motions which degrades the robotŐs performance during a game. This paper presents an implementation of the policy gradient machine learning algorithm that searches for a parameterized walk while optimizing for both speed and stability. The experiments were performed on the Sony Aibo ERS-7 robot platform. The resulting gait is reasonably fast and considerably more stable compared to AustinVillaŐs fast gait.
Stronger, Daniel and Peter Stone. "Polynomial Regression with Automated Degree: A Function Approximator for Autonomous Agents." The University of Texas at Austin, Department of Computer Sciences. AI Technical Report 06-329. May 10, 2006. 12 pages.
In order for an autonomous agent to behave robustly in a variety of environments, it must have the ability to learn approximations to many different functions. The function approximator used by such an agent is subject to a number of constraints that may not apply in a traditional supervised learning setting. Many different function approximators exist and are appropriate for different problems. This paper proposes a set of criteria for function approximators for autonomous agents. Additionally, for those problems on which polynomial regression is a candidate technique, the paper presents an enhancement that meets these criteria. In particular, using polynomial regression typically requires a manual choice of the polynomial's degree, trading off between function accuracy and computational and memory efficiency. Polynomial Regression with Automated Degree (PRAD) is a novel function approximation method that uses training data to automatically identify an appropriate degree for the polynomial. PRAD is fully implemented. Empirical tests demonstrate its ability to efficiently and accurately approximate both a wide variety of synthetic functions and real-world data gathered by a mobile robot.
Wildstrom, Jonathan, Peter Stone, Emmett Witchel, and Mike Dahlin. "Adapting to Workload Changes Through On-The-Fly Reconfiguration." The University of Texas at Austin, Department of Computer Sciences. AI Technical Report 06-330. May 31, 2006. 15 pages.
High-end servers that can be partitioned into logical subsystems and repartitioned on the fly are now becoming available. This development raises the possibility of reconfiguring distributed systems online to optimize for dynamically changing workloads. This paper presents one approach to solving this online reconfiguration problem. In particular, we learn to identify, from only low-level system statistics, which of a set of possible configurations will lead to better performance under the current unknown workload. This approach requires no instrumentation of the system's middleware or operating systems. We introduce an agent that is able to learn this model and use it to switch configurations online as the workload varies. Our agent is fully implemented and tested on a publically available multi-machine, multi-process distributed system (the online transaction processing benchmark TPC-W). We demonstrate that our adaptive configuration is able to outperform any single fixed configuration in the set over a variety of workloads, including gradual changes and abrupt workload spikes.
Clark, Peter, Bruce Porter, and Don Batory. "A Compositional Approach to Representing Planning Operators." The University of Texas at Austin, Department of Computer Sciences. AI Technical Report 06-331. April 19, 2006. 14 pages.
AI has frequently been criticized for being `stuck in the microworld' because of the common inability of AI systems to cope with the complexity of real domains. Often, adding details removes regularity, transforming a representation from a few simple structures to a large, unwieldy collection of specialized ones. This paper addresses this problem in the context of representing planning operators (domain-specific knowledge about the effects of actions in a domain) for use by AI planning systems. We present a novel approach in which domain-specific operators are represented as a composition of general components, and show that the problem of manually building a detailed set of operators can be avoided by constructing them from a small number of such components instead. Each component encapsulates information about a domain feature that might be modeled, and each may contribute to several operators. Moreover, we describe how the choice of what to model and what to ignore in a domain can then be easily varied, simply by controlling which components are used. Finally, we show how operator sets built in this way can be used by planning algorithms.
Clark, Peter and Bruce Porter. "Building Domain Representations from Components." The University of Texas at Austin, Department of Computer Sciences. AI Technical Report 06-332. April 19, 2006. 15 pages.
A major cause of the knowledge-engineering bottleneck is the inability to trans fer representational fragments from one knowledge base to another due to the idiosyncratic nature of domain-specific representations. In this paper, we show that representations can be built automatically by composing abstract, reusable components. Moreover, we describe how representations of specific situations, that arise during problem solving, can be assembled `on demand', guided by a query for a particular piece of information. Our work integrates ideas from dynamic memory, conceptual graph theory, compositional modeling and graph unification.
Monroy, German A., Kenneth O. Stanley, and Risto Miikkulainen. "Coevolution of Neural Networks Using a Layered Pareto Archive." The University of Texas at Austin, Department of Computer Sciences. AI Technical Report 06-333. August 2005. 63 pages.
Evolutionary computation (EC) can be used to discover good game playing strategies with respect to a fixed group of existing opponents. Competitive coevolution (CC) is a generalization of EC in which the opponents also evolve, making it possible to find strategies for games in which no good players already exist. Since the performance criterion in CC changes over generations, a Coevolutionary Memory (CM) of good opponents is required to avoid regress. The Layered Pareto Coevolution Archive (LAPCA) was recently proposed as an effective CM that guarantees monotonic progress under certain assumptions. The main limitation of LAPCA is that it requires numerous game evaluations because it has to determine Pareto-dominance relations between players. The Hall of Fame (HOF), consisting of previous generation champions, is an easier CM to implement and needs no extra evaluations besides those required by evolution. While the LAPCA has only been demonstrated in artificial numeric games, the HOF has been applied to real world problems such as the coevolution of neural networks. This thesis makes three main contributions. First, a technique is developed that interfaces the LAPCA algorithm with NeuroEvolution of Augmenting Topologies (NEAT), which has been shown to be an efficient method of neuroevolution in game playing domains. The technique is shown to keep the total number of evaluations in the order of those required by NEAT, making it applicable to practical domains. Second, the behavior of LAPCA is analyzed for the first time in a complex game playing domain: evolving neural network players for the game of Pong. Third, although LAPCA and HOF perform equally well in this domain, LAPCA is shown to require significantly less space than the HOF. Therefore, combining NEAT and LAPCA is found to be an effective approach to coevolution; the main task for the future is to test it in domains that are more susceptible to forgetting than Pong, where it can potentially lead to improved performance as well.
Bryant, Bobby. "DISSERTATION." The University of Texas at Austin, Department of Computer Sciences. AI Technical Report 06-334. .
Yang, Stewart, Jianping Song, Harish Rajamani, Taewon Cho, Yin Zhang, and Raymond Mooney. "Fast and Effective Worm Fingerprinting via Machine Learning." The University of Texas at Austin, Department of Computer Sciences. AI Technical Report 06-335. August 8, 2006. 10 pages.
As Internet worms become ever faster and more sophisticated, it is important to be able to extract worm signatures in an accurate and timely manner. In this paper, we apply machine learning to automatically fingerprint polymorphic worms, which are able to change their appearance across every instance. Using real Internet traces and synthetic polymorphic worms, we evaluated the performance of several advanced machine learning algorithms, including naive Bayes, decision-tree induction, rule learning, and support vector machines. The results are very promising. Compared with Polygraph, the state of the art in polymorphic worm fingerprinting, several machine learning algorithms are able to generate more accurate signatures, tolerate more noise in the training data, and require much shorter training time. These results open the possibility of applying machine learning to build a fast and accurate online worm fingerprinting system.
Whiteson, Shimon and Daniel Whiteson. "Stochastic Optimization for Collision Selection in High Energy Physics." The University of Texas at Austin, Department of Computer Sciences. AI Technical Report 06-336. August 10, 2006. 6 pages.
The underlying structure of matter can be deeply probed via precision measurements of the mass of the top quark, the most massive observed fundamental particle. Top quarks can be produced and studied only in collisions at high energy particle accelerators. Most collisions, however, do not produce top quarks; making precise measurements requires culling these collisions into a sample that is rich in collisions producing top quarks (signal) and spare in collisions producing other particles (background). Collision selection is typically performed with heuristics or supervised learning methods. However, such approaches are suboptimal because they assume that the selector with the highest classification accuracy will yield a mass measurement with the smallest statistical uncertainty. In practice, however, the mass measurement is more sensitive to some backgrounds than others. Hence, this paper presents a new approach that uses stochastic optimization techniques to directly search for selectors that minimize statistical uncertainty in the top quark mass measurement. Empirical results confirm that stochastically optimized selectors have much smaller uncertainty. This new approach contributes substantially to our knowledge of the top quark's mass, as the new selectors are currently in use selecting real collisions.
Stone, Peter, Peggy Fidelman, Nate Kohl, Gregory Kuhlmann, Tekin Mericli, Mohan Sridharan, and Shao-en Yu. "The UT Austin Villa 2006 RoboCup Four-Legged Team." The University of Texas at Austin, Department of Computer Sciences. AI Technical Report 06-337. December 18, 2006. 13 pages.
The UT Austin Villa Four-Legged Team for RoboCup 2006 was a fourth-time entry in the ongoing series of RoboCup legged league competitions. The team development began in mid-January of 2003 without any prior familiarity with the Aibos. After entering a fairly non-competitive team in RoboCup 2003, the team made several important advances. By the July 2004 competition that took place in Lisbon, Portugal, it was one of the top few teams. After those first two years of intense development, the team's third and fourth years were devoted more to spin-off research than to development related to the competition. Building off of the team's previous three technical reports, this report details the changes made to the team between RoboCup 2005 and RoboCup 2006 in Bremen. Taken together, this and the previous technical reports provide the history and details of the UT Austin Villa RoboCup Four-Legged team and the associated research in our lab.
Questions to trcenter@cs.utexas.edu