Browsing by Author "Dr. Dennis Bahler, Committee Member"

Now showing 1 - 4 of 4

Building an Essential Gene Classification Framework
(2006-01-05) Saha, Soma; Dr. Dennis Bahler, Committee Member; Dr. Xiaosong Ma, Committee Member; Dr. Steffen Heber, Committee Chair
The analysis of gene deletions is a fundamental approach for investigating gene function. We applied machine learning techniques to predict phenotypic effects of gene deletions in yeast. We created a dataset containing features that potentially have predictive power and then used feature processing techniques to improve the dataset and identify features that are important for our classification problem. We evaluated four different classification algorithms, K-Nearest Neighbors, Support Vector Machine, Decision Tree, and Random Forest, with respect to this problem. We used our framework to complement the set of experimentally determined essential yeast genes produced by the Saccharomyces Genome Deletion Project and produce more than 2000 annotations for genes that might cause morphological alterations in yeast.
On Learning of Ceteris Paribus Preference Theories
(2007-04-06) Sachdev, Manish Prakash; Dr. Dennis Bahler, Committee Member; Dr. Munindar P. Singh, Committee Member; Dr. Jon Doyle, Committee Chair
The problem of preference elicitation has been of interest for a long time. While traditional methods of asking a set of relevant questions are still useful, the availability of user-preference data from the web has led to substantial attention to the notion of preference mining. In this thesis, we consider the problem of learning logical preference theories that express preference orderings over alternatives. We present learning algorithms which accept as input a set of comparisons between pairs of complete descriptions of world states. Our first algorithm, that performs exact learning, accepts the complete set of preference orderings for a theory and generates a theory which provides the same ordering of states as the input. This process can require looking at an exponential number of data points. We then look at more realistic approximation algorithms and analyze the complexity of the learning problem under the framework of Probably Approximately Correct (PAC) learning. We then describe approximation algorithms for learning high-level summaries of the underlying theory.
Plan Recognition as Candidate Space Search
(2007-03-06) Kumaran, Vikram; Dr. Dennis Bahler, Committee Member; Dr. Michael Young, Committee Chair; Dr James Lester, Committee Member
Effective human computer interaction is enhanced by a machine's ability to make educated guesses about the intention of its user. In our research, we have developed a novel plan recognition algorithm — based on plan space search planners — to recognize plans given a limited set of observed actions. Our focus in this research is towards accurately picking possible plans and not towards disambiguation or building plan libraries and therefore we complement other advances in this field, namely probability based recognition and other plan library based recognition systems. Along with the ability to recognize overall goal of an agent our algorithm also allows us to make local predictions, a feature absent in most of the other system.
A Web Services Approach to Generating and Using Plans in Configurable Execution Environments
(2006-03-01) Vernieri, Thomas Michael; Dr. R. Michael Young, Committee Chair; Dr. Dennis Bahler, Committee Member; Dr. James Lester, Committee Member
The computational scope of artificial intelligence in games has traditionally been limited by the processing requirements of the game engine's graphics and physics components. The emerging genre of interactive narrative typically relies upon AI planning systems that perform computation too demanding to integrate into commercial games. This thesis describes Zocalo, a collection of service-oriented applications, in which a planning Web service generates interactive storylines for story-based games and interactive narratives. The interfaces of the planning services allow for usage scenarios ranging from simple to complex. We describe the use of planning services both for run-time construction of narrative plans and for design-time iterative specification of game contents. Zocalo facilitates the execution of plans in commercial game engines with only small modifications to the original games. It provides for the execution of story plans in numerous situations, automatically adjusting the state of the game's environment so that it is compatible with the beginning of the story. The plan execution functionality of Zocalo is intended for gaming environments but can also be applied to other applications in need of narratives.