摘要 :
In designing for a system's lifecycle considerations, longterm energy needs often become an important limiting factor. Shifting from conventional energy sources (e.g. fossil fuels)toward renewable sources (e.g. wind and solar) has...
展开
In designing for a system's lifecycle considerations, longterm energy needs often become an important limiting factor. Shifting from conventional energy sources (e.g. fossil fuels)toward renewable sources (e.g. wind and solar) has become a popular means for focusing on the lifecycle of large-scale systems like automobiles and the national electrical grid. This same shift in small, low-power systems such as sensors has the additional advantage of potentially increasing the operational life of the systems.
收起
摘要 :
Effective and proactive decisions about intelligence gathering depend on accurate models of an adversary. Specifically, such models need to accurately reflect the cause-and-effect dependencies within the systemic behavior of the a...
展开
Effective and proactive decisions about intelligence gathering depend on accurate models of an adversary. Specifically, such models need to accurately reflect the cause-and-effect dependencies within the systemic behavior of the adversary. Such models can be created based entirely on the knowledge of experts, or they can be created or augmented based on the analysis of data. However, creating causal models from data will require advances in the fundamental science and technology of discovering causal knowledge. Our project focused on creating such advances. Specifically, we focused on automating the application of quasi-experimental designs, a set of manual analysis techniques developed by social scientists, economists, and medical researchers over the past four decades. Quasi-experimental designs (QEDs) are templates for causal discovery from observational (non-experimental) data. QEDs identify naturally occurring experiments that support inferences about causal dependencies within large bodies of observational data. Our work has shown that many potential designs exist for realistic tasks that those designs can increase the accuracy with which causal inferences can be made from small amounts of data, and that such designs can be automatically identified. This lays the groundwork for powerful tools with which analysts can examine observational data of complex organizations and system to improve their causal understanding of those systems.
收起
摘要 :
The Omaha District (District) of the U.S. Army Corps of Engineers (Corps) is implementing a Water Quality Management Program (WQMP) as part of the operation and maintenance activities associated with managing the Corps civil works...
展开
The Omaha District (District) of the U.S. Army Corps of Engineers (Corps) is implementing a Water Quality Management Program (WQMP) as part of the operation and maintenance activities associated with managing the Corps civil works projects in the District
收起
摘要 :
A persistent goal of research in artificial intelligence has been to enable learning and reasoning with probabilistic models in complex domains. Much of this work has been directed toward systems that complement, rather than repla...
展开
A persistent goal of research in artificial intelligence has been to enable learning and reasoning with probabilistic models in complex domains. Much of this work has been directed toward systems that complement, rather than replace, human abilities and knowledge. Models that fuse engineering knowledge (knowledge from human sources) with learned information (information gained algorithmically) can take advantage of the strengths of both approaches, yielding more accurate predictions. A particularly fruitful area for this research is improving our understanding of emergent behavior, specifically, how connectivity among individual units of a system affects global behavior. The Knowledge Discovery Laboratory (KDL) seeks to apply a growing understanding of emergent behavior to the design of learning and reasoning systems.
收起
摘要 :
The introduction of active learning exercises into a traditional lecture has been shown to improve students' learning. Hands-on learning opportunities in labs and projects provide are additional tools in the active learning toolbo...
展开
The introduction of active learning exercises into a traditional lecture has been shown to improve students' learning. Hands-on learning opportunities in labs and projects provide are additional tools in the active learning toolbox. This paper presents a series of innovative hands-on active learning activities for mechanics of materials topics. These activities are based on a 'Methodology for Developing Hands-on Active Learning Activities,' a systematic approach for efficient and effective activity development. The activities are being evaluated at three institutions of higher learning: Austin Community College, the U.S. Air Force Academy, and the University of Texas at Austin. Seven of the 28 activities have been evaluated to date. Evaluation consists of a variety of measures, including student opinion surveys, focus groups, pre-post activity quizzes, exam questions, and a concept inventory. In addition, information on demography and student learning styles was collected, and Myers-Briggs personality type was assessed. This information was correlated to the student evaluation measures. Data from over 150 students are summarized and discussed. The results show that, in general, students are excited about doing hands-on activities during lectures, and they believe that the activities enhance their learning. While these general findings exist, students' learning style, personality type, and perception of performance in the class all influence their opinions of the activities and will be measured further in future activity development and evaluation.
收起
摘要 :
Bias/variance analysis is a useful tool for investigating the performance of machine learning algorithms. Conventional analysis decomposes loss into errors due to aspects of the learning process, but in relational domains, the inf...
展开
Bias/variance analysis is a useful tool for investigating the performance of machine learning algorithms. Conventional analysis decomposes loss into errors due to aspects of the learning process, but in relational domains, the inference process introduces an additional source of error. Collective inference techniques introduce additional error both through the use of approximate inference algorithms and through variation in the availability of test set information. To date, the impact of inference error on model performance has not been investigated. In this paper, we propose a new bias/variance framework that decomposes loss into errors due to both the learning and inference process. We evaluate performance of three relational models and show that (1) inference can be a significant source of error, and (2) the models exhibit different types of errors as data characteristics are varied.
收起
摘要 :
Statistical analysis of relational data is a fundamental and novel problem in machine learning and data mining. Such analysis constructs useful statistical models from data about complex relationships among people, places, things,...
展开
Statistical analysis of relational data is a fundamental and novel problem in machine learning and data mining. Such analysis constructs useful statistical models from data about complex relationships among people, places, things, and events. Supported by this research contract, we uncovered fundamental challenges of statistical learning and inference in relational data, we designed and implemented new languages for expressing deterministic and probabilistic dependencies in such data, we developed new algorithms for learning probabilistic models, we implemented an open-source system for knowledge discovery in relational data containing over 40,000 lines of code that has been downloaded more than 1000 times, and we evaluated the utility of those algorithms by undertaking large and realistic applications.
收起
摘要 :
This research developed techniques for data analysis of multi-agent systems. The research focused on how to analyze relational data that represent sets of interconnected agents, resources, and locations, as well as the attributes ...
展开
This research developed techniques for data analysis of multi-agent systems. The research focused on how to analyze relational data that represent sets of interconnected agents, resources, and locations, as well as the attributes of these objects. Results of the research include new algorithms for knowledge discovery, fundamental discoveries about the challenges and opportunities of constructing statistical models in relational data, software prototypes of the algorithms developed, and a simulation environment for evaluating the techniques.
收起
