The unprecedented growth in numbers of children playing computer games has stimulated discussion ... more The unprecedented growth in numbers of children playing computer games has stimulated discussion and research regarding what, if any, educational value these games have for teaching and learning. The research on this topic has primarily focused on children as players of computer games rather than builders/constructors of computer games. Recently, several game companies, such as BioWare Corp. and Bethesda Softworks, have released game story creation tools to the public, along with their games. However, a major obstacle to using these commercial tools is the level of programming experience required to create interactive game stories. In this paper, we demonstrate that a commercial game story construction tool, BioWare Corp.'s Aurora Toolset, can be augmented by our new tool, ScriptEase, to enable students in two grade ten English classes to successfully construct interactive game stories. We present evidence that describes the relationship between interactive story authoring and traditional story authoring, along with a series of factors that can potentially affect success at these activities: gender, creativity, intellectual ability, previous experiences with programming, time playing computer games, and time spent online. Results indicate that students can successfully construct sophisticated interactive stories with very little training. The results also show no gender differences in the quality of these interactive stories, regardless of programming experience or the amount of time per week playing computer games or participating in general online activities, although a subset of female students did show a slightly higher level of performance on interactive story authoring. In the educational context of this study, we show that ScriptEase provides an easy-to-use tool for interactive story authoring in a constructionist learning environment.
national conference on artificial intelligence, Jul 28, 2002
The time and space needs of an A* search are strongly influenced by the quality of the heuristic ... more The time and space needs of an A* search are strongly influenced by the quality of the heuristic evaluation function. Usually there is a trade-off since better heuristics may require more time and/or space to evaluate. Multiple sequence alignment is an important application for single-agent search. The traditional heuristic uses multiple pairwise alignments that require relatively little space. Three-way alignments produce better heuristics, but they are not used in practice due to the large space requirements. This paper presents a memory-efficient way to represent three-way heuristics as an octree. The required portions of the octree are computed on demand. The octree-supported three-way heuristics result in such a substantial reduction to the size of the A* open list that they offset the additional space and time requirements for the three-way alignments. The resulting multiple sequence alignments are both faster and use less memory than using A* with traditional pairwise heuristics.
A thesis submitted to the Faculty of Graduate Studies and Research in partial ful llment of the r... more A thesis submitted to the Faculty of Graduate Studies and Research in partial ful llment of the requirements for the degree of Master of Science.
Multi-agent decision problems can often be formulated as extensive-form games. We focus on imperf... more Multi-agent decision problems can often be formulated as extensive-form games. We focus on imperfect information extensive-form games in which one or more actions at many decision points have an associated continuous or manyvalued parameter. A stock trading agent, in addition to deciding whether to buy or not, must decide how much to buy. In no-limit poker, in addition to selecting a probability for each action, the agent must decide how much to bet for each betting action. Selecting values for these parameters makes these games extremely large. Two-player no-limit Texas Hold’em poker with stacks of 500 big blinds has approximately 10 71 states, which is more than 10 50 times more states than twoplayer limit Texas Hold’em. The main contribution of this paper is a technique that abstracts a game’s action space by selecting one, or a small number, of the many values for each parameter. We show that strategies computed using this new algorithm for no-limit Leduc poker exhibit significa...
In extensive-form games with a large number of actions, careful abstraction of the action space i... more In extensive-form games with a large number of actions, careful abstraction of the action space is critically important to performance. In this paper we extend previous work on action abstraction using no-limit poker games as our test domains. We show that in such games it is no longer necessary to choose, a priori, one specific range of possible bet sizes. We introduce an algorithm that adjusts the range of bet sizes considered for each bet individually in an iterative fashion. This flexibility results in a substantially improved game value in no-limit Leduc poker. When applied to no-limit Texas Hold’em our algorithm produces an action abstraction that is about one third the size of a state of the art hand-crafted action abstraction, yet has a better overall game value.
In this work we present a uniform behavioral temporal object model which includes a rich and exte... more In this work we present a uniform behavioral temporal object model which includes a rich and extensible set of types and behaviors to support various notions of time. Our temporal model supports the continuous and discrete domains of time. It also supports various speci cations of time namely, time instants, time intervals, and time spans. Two issues which deal with temporal information that frequently arise in real-world applications are the need to consistently store and operate on temporal information that is comprised of di erent granularities, and the ability to represent indeterminate temporal information. Temporal indeterminacy arises when we know for certain that an event did occur, but exactly when it occured is unknown. Our model provides a uniform framework that supports di erent granularities of time and temporal indeterminacy. We show how di erent granularities can be converted to each other, and how this leads to temporal indeterminacy. We further show how continuous a...
We describe the design of a multimedia database for a distributed news-on-demand multimedia infor... more We describe the design of a multimedia database for a distributed news-on-demand multimedia information system. News–on–demand is an application that utilizes broadband network services to deliver news articles to subscribers in the form of multimedia documents. Different news providers insert articles into the database, which is then accessed by remote users over a broadband network. Multimedia documents are composite objects where the component objects have spatial and temporal relationships which need to be captured in the database. Modeling of multimedia documents involves three issues: (1) modeling of individual document components (i.e., monomedia objects such as text, images, etc), (2) modeling of the document structure, and (3) modeling of the presentation structure. We take an object-oriented approach to dealing with these issues. Within (1), our research has so far concentrated on text. We use an annotation based scheme where entire text is stored as a single document and ...
Video modeling has become a topic of increasing interest in the area of multimedia research. One ... more Video modeling has become a topic of increasing interest in the area of multimedia research. One of the key aspects in the video medium is the temporal relationship between video frames. In this report1 we propose a tree-based model for specifying spatial and temporal semantics of video data. Our focus here is on the temporal issues. We present a unique way of integrating our video model into an objectbase management system which has rich multimedia temporal operations. We further show how temporal histories are used to model video data. Using histories to model video data is both simple and natural. It also can lead to a uniform behavioral model. A user can then explore the video objectbase using object-oriented techniques. Such a seamless integration gives a uniform interface to end users. The integrated video objectbase management system supports a broad range of temporal queries and is extensible, thus allowing the easy incorporation of new features into the system.
Motivation: Modern sequencing technology now permits the sequencing of entire genomes, leading to... more Motivation: Modern sequencing technology now permits the sequencing of entire genomes, leading to thousands of new gene sequences in need of detailed annotation. It is too time consuming to predict the properties of each protein sequence manually and to organize the results of many prediction tools by hand. The prediction process must be automated so the predictions can be automatically organized, but the predictions must also be transparent. That is, the rationale for each prediction should be easily examinable by anyone that wishes to use the prediction. Results: Proteome Analyst (PA) is a web-based system for predicting the properties of each protein in a proteome. PA has three interesting features. First, it provides a single web-based tool that allows the user to select a wide range of analytic tools and automatically apply them to each protein in a proteome. In essence, PA provides one-stop automatic high-throughput analysis. Second, PA has the ability to explain its predictio...
Declarative query languages are an important feature of database management systems and have play... more Declarative query languages are an important feature of database management systems and have played an important role in their success. As database management technology enters the multimedia information system area, the availability of special-purpose query languages for multimedia applications will be equally important. In this report 1 , we survey multimedia query languages and query models. Particularly, we look at those systems from the point of view of well-deened queries, fuzzy queries, visual queries, and query presentations. Several research issues, such as generic multimedia query languages, incremental queries, fuzzy queries, spatio-temporal queries, feature storage and organization, are also identiied. In our opinion these are vital issues for the success and development of a multimedia query language.
Object oriented computing is in uencing many areas of computer science including software en gine... more Object oriented computing is in uencing many areas of computer science including software en gineering user interfaces operating systems programming languages and database systems The appeal of object orientation is attributed to its higher levels of abstraction for modeling real world concepts its support for incremental development and its potential for interoperability Despite many advances object oriented computing is still in its infancy and a universally acceptable def inition of an object oriented data model is virtually nonexistent although some standardization e orts are underway This report presents the TIGUKAT object model de nition that is the result of an investigation of object oriented modeling features which are common among earlier proposals along with some distinctive qualities that extend the power and expressibility of this model beyond others The literature recognizes two perspectives of an object model the structural view and the behavioral view Most object ori...
Games are used to evaluate and advance Multiagent and Artificial Intelligence techniques. Most of... more Games are used to evaluate and advance Multiagent and Artificial Intelligence techniques. Most of these games are deterministic with perfect information (e.g. Chess and Checkers). A deterministic game has no chance element and in a perfect information game, all information is visible to all players. However, many real-world scenarios with competing agents are stochastic (non-deterministic) with imperfect information. For two-player zero-sum perfect recall games, a recent technique called Counterfactual Regret Minimization (CFR) computes strategies that are provably convergent to an e-Nash equilibrium. A Nash equilibrium strategy is useful in two-player games since it maximizes its utility against a worst-case opponent. However, for multiplayer (three or more player) games, we lose all theoretical guarantees for CFR. However, we believe that CFR-generated agents may perform well in multiplayer games. To test this hypothesis, we used this technique to create several 3-player limit Tex...
Naive Bayes classifiers, a popular tool for predicting the labels of query instances, are typical... more Naive Bayes classifiers, a popular tool for predicting the labels of query instances, are typically learned from a training set. However, since many training sets contain noisy data, a classifier user may be reluctant to blindly trust a predicted label. We present a novel graphical explanation facility for Naive Bayes classifiers that serves three purposes. First, it transparently explains the reasoning used by the classifier to foster user confidence in the prediction. Second, it enhances the user's understanding of the complex relationships between the features and the labels. Third, it can help the user to identify suspicious training data. We demonstrate these ideas in the context of our implemented web-based system, which uses examples from molecular biology.
Workstation environments have been in use for more than a decade now. Although a network of works... more Workstation environments have been in use for more than a decade now. Although a network of workstations together represents a large amount of aggregate computing power, single users often cannot utilize these resources for their applications. E terprise is a programming environment for designing, coding, debugging, testing, monitoring, profiling and executing programs in a distributed hardware environment. Enterprise code looks like familiar sequential code; the parallelism is expressed graphically. The system automatically inserts the code necessary to handle communication, synchronization and fault tolerance, allowing the rapid construction of correct distributed programs. Enterprise programs run on a network of computers, absorbing the idle cycles on machines. The system supports load balancing, limited process migration, and dynamic distribution of work in environments with changing resource utilization. E terprise offers a costeffective method for increasing the productivity o...
In this paper we consider various spatial relationships that are of general interest for retrievi... more In this paper we consider various spatial relationships that are of general interest for retrieving data from multimedia databases. We present a uni ed representation of spatial objects for both topological and directional relations. Such a representation is based on Allen's temporal interval algebra. We also present a set of spatial inference rules, which allow us to make heterogeneous spatial relation deductions from existing directional and topological relations. For example, if we know A north of B, B overlap with C, and C north of D, then we deriveA above D. Since all the rules are propositional Horn clauses, they can be easily integrated into any multimedia database by using either a simple inference engine or a lookup table.
The Enterprise programming environment supports the development of applications that run concurre... more The Enterprise programming environment supports the development of applications that run concurrently on a network of workstations. This paper describes the object-oriented components of Enterprise, implemented in Smalltalk-80, and their seamless integration with the procedural components, implemented in C. The object-oriented user-interface supports a new anthropomorphic model for parallel computation that eliminates much of the perceived complexity of parallel programs. The object-oriented animation component implements a new animation architecture that supports synchronous and asynchronous events. This allows a user to view the dynamic interactions of the parallel components of a distributed application to simplify performance monitoring and debugging. The Enterprise experience highlights the strengths of object-oriented methodologies both for expressing user models and for implementing related components.
In large extensive form games with imperfect information, Counterfactual Regret Minimization (CFR... more In large extensive form games with imperfect information, Counterfactual Regret Minimization (CFR) is a popular, iterative algorithm for computing approximate Nash equilibria. While the base algorithm performs a full tree traversal on each iteration, Monte Carlo CFR (MCCFR) reduces the per iteration time cost by traversing just a sampled portion of the tree. On the other hand, MCCFR's sampled values introduce variance, and the effects of this variance were previously unknown. In this paper, we generalize MCCFR by considering any generic estimator of the sought values. We show that any choice of an estimator can be used to probabilistically minimize regret, provided the estimator is bounded and unbiased. In addition, we relate the variance of the estimator to the convergence rate of an algorithm that calculates regret directly from the estimator. We demonstrate the application of our analysis by defining a new bounded, unbiased estimator with empirically lower variance than MCCFR...
Modeling moving objects has become a topic of increasing interest in the area of video databases.... more Modeling moving objects has become a topic of increasing interest in the area of video databases. Two key aspects of such modeling are object spatial and temporal relationships. In this paper we introduce an innovative way to represent the trajectory of single moving object and the relative spatio-temporal relations between multiple moving objects. The representation supports a rich set of spatial topological and directional relations. It also supports both quantitative and qualitative user queries about moving objects. Algorithms for matching trajectories and spatio-temporal relations of moving objects are designed to facilitate query processing. These algorithms can handle both exact and similarity matches. We also discuss the integration of our moving object model, based on a video model, in an object-oriented system. Some query examples are provided to further validate the expressiveness of our model.
A "first" implementation of the Modular Smalltalk object-oriented programming language ... more A "first" implementation of the Modular Smalltalk object-oriented programming language is presented. The implementation includes an object-oriented parser, object-oriented representation for code fragments and an object-oriented C-code generator, all implemented in Smalltalk-80. This implementation validates two of the five design principles of the Modular Smalltalk language and provides a vehicle for validating the other three design principles. The macro-based C-code generator is easily adaptable to generating production code in other languages like assembler. In addition, the generation technique applies to source languages other than Modular Smalltalk. The implementation includes an efficient method dispatch based on new extensions to incremental cache table coloring.
The unprecedented growth in numbers of children playing computer games has stimulated discussion ... more The unprecedented growth in numbers of children playing computer games has stimulated discussion and research regarding what, if any, educational value these games have for teaching and learning. The research on this topic has primarily focused on children as players of computer games rather than builders/constructors of computer games. Recently, several game companies, such as BioWare Corp. and Bethesda Softworks, have released game story creation tools to the public, along with their games. However, a major obstacle to using these commercial tools is the level of programming experience required to create interactive game stories. In this paper, we demonstrate that a commercial game story construction tool, BioWare Corp.'s Aurora Toolset, can be augmented by our new tool, ScriptEase, to enable students in two grade ten English classes to successfully construct interactive game stories. We present evidence that describes the relationship between interactive story authoring and traditional story authoring, along with a series of factors that can potentially affect success at these activities: gender, creativity, intellectual ability, previous experiences with programming, time playing computer games, and time spent online. Results indicate that students can successfully construct sophisticated interactive stories with very little training. The results also show no gender differences in the quality of these interactive stories, regardless of programming experience or the amount of time per week playing computer games or participating in general online activities, although a subset of female students did show a slightly higher level of performance on interactive story authoring. In the educational context of this study, we show that ScriptEase provides an easy-to-use tool for interactive story authoring in a constructionist learning environment.
national conference on artificial intelligence, Jul 28, 2002
The time and space needs of an A* search are strongly influenced by the quality of the heuristic ... more The time and space needs of an A* search are strongly influenced by the quality of the heuristic evaluation function. Usually there is a trade-off since better heuristics may require more time and/or space to evaluate. Multiple sequence alignment is an important application for single-agent search. The traditional heuristic uses multiple pairwise alignments that require relatively little space. Three-way alignments produce better heuristics, but they are not used in practice due to the large space requirements. This paper presents a memory-efficient way to represent three-way heuristics as an octree. The required portions of the octree are computed on demand. The octree-supported three-way heuristics result in such a substantial reduction to the size of the A* open list that they offset the additional space and time requirements for the three-way alignments. The resulting multiple sequence alignments are both faster and use less memory than using A* with traditional pairwise heuristics.
A thesis submitted to the Faculty of Graduate Studies and Research in partial ful llment of the r... more A thesis submitted to the Faculty of Graduate Studies and Research in partial ful llment of the requirements for the degree of Master of Science.
Multi-agent decision problems can often be formulated as extensive-form games. We focus on imperf... more Multi-agent decision problems can often be formulated as extensive-form games. We focus on imperfect information extensive-form games in which one or more actions at many decision points have an associated continuous or manyvalued parameter. A stock trading agent, in addition to deciding whether to buy or not, must decide how much to buy. In no-limit poker, in addition to selecting a probability for each action, the agent must decide how much to bet for each betting action. Selecting values for these parameters makes these games extremely large. Two-player no-limit Texas Hold’em poker with stacks of 500 big blinds has approximately 10 71 states, which is more than 10 50 times more states than twoplayer limit Texas Hold’em. The main contribution of this paper is a technique that abstracts a game’s action space by selecting one, or a small number, of the many values for each parameter. We show that strategies computed using this new algorithm for no-limit Leduc poker exhibit significa...
In extensive-form games with a large number of actions, careful abstraction of the action space i... more In extensive-form games with a large number of actions, careful abstraction of the action space is critically important to performance. In this paper we extend previous work on action abstraction using no-limit poker games as our test domains. We show that in such games it is no longer necessary to choose, a priori, one specific range of possible bet sizes. We introduce an algorithm that adjusts the range of bet sizes considered for each bet individually in an iterative fashion. This flexibility results in a substantially improved game value in no-limit Leduc poker. When applied to no-limit Texas Hold’em our algorithm produces an action abstraction that is about one third the size of a state of the art hand-crafted action abstraction, yet has a better overall game value.
In this work we present a uniform behavioral temporal object model which includes a rich and exte... more In this work we present a uniform behavioral temporal object model which includes a rich and extensible set of types and behaviors to support various notions of time. Our temporal model supports the continuous and discrete domains of time. It also supports various speci cations of time namely, time instants, time intervals, and time spans. Two issues which deal with temporal information that frequently arise in real-world applications are the need to consistently store and operate on temporal information that is comprised of di erent granularities, and the ability to represent indeterminate temporal information. Temporal indeterminacy arises when we know for certain that an event did occur, but exactly when it occured is unknown. Our model provides a uniform framework that supports di erent granularities of time and temporal indeterminacy. We show how di erent granularities can be converted to each other, and how this leads to temporal indeterminacy. We further show how continuous a...
We describe the design of a multimedia database for a distributed news-on-demand multimedia infor... more We describe the design of a multimedia database for a distributed news-on-demand multimedia information system. News–on–demand is an application that utilizes broadband network services to deliver news articles to subscribers in the form of multimedia documents. Different news providers insert articles into the database, which is then accessed by remote users over a broadband network. Multimedia documents are composite objects where the component objects have spatial and temporal relationships which need to be captured in the database. Modeling of multimedia documents involves three issues: (1) modeling of individual document components (i.e., monomedia objects such as text, images, etc), (2) modeling of the document structure, and (3) modeling of the presentation structure. We take an object-oriented approach to dealing with these issues. Within (1), our research has so far concentrated on text. We use an annotation based scheme where entire text is stored as a single document and ...
Video modeling has become a topic of increasing interest in the area of multimedia research. One ... more Video modeling has become a topic of increasing interest in the area of multimedia research. One of the key aspects in the video medium is the temporal relationship between video frames. In this report1 we propose a tree-based model for specifying spatial and temporal semantics of video data. Our focus here is on the temporal issues. We present a unique way of integrating our video model into an objectbase management system which has rich multimedia temporal operations. We further show how temporal histories are used to model video data. Using histories to model video data is both simple and natural. It also can lead to a uniform behavioral model. A user can then explore the video objectbase using object-oriented techniques. Such a seamless integration gives a uniform interface to end users. The integrated video objectbase management system supports a broad range of temporal queries and is extensible, thus allowing the easy incorporation of new features into the system.
Motivation: Modern sequencing technology now permits the sequencing of entire genomes, leading to... more Motivation: Modern sequencing technology now permits the sequencing of entire genomes, leading to thousands of new gene sequences in need of detailed annotation. It is too time consuming to predict the properties of each protein sequence manually and to organize the results of many prediction tools by hand. The prediction process must be automated so the predictions can be automatically organized, but the predictions must also be transparent. That is, the rationale for each prediction should be easily examinable by anyone that wishes to use the prediction. Results: Proteome Analyst (PA) is a web-based system for predicting the properties of each protein in a proteome. PA has three interesting features. First, it provides a single web-based tool that allows the user to select a wide range of analytic tools and automatically apply them to each protein in a proteome. In essence, PA provides one-stop automatic high-throughput analysis. Second, PA has the ability to explain its predictio...
Declarative query languages are an important feature of database management systems and have play... more Declarative query languages are an important feature of database management systems and have played an important role in their success. As database management technology enters the multimedia information system area, the availability of special-purpose query languages for multimedia applications will be equally important. In this report 1 , we survey multimedia query languages and query models. Particularly, we look at those systems from the point of view of well-deened queries, fuzzy queries, visual queries, and query presentations. Several research issues, such as generic multimedia query languages, incremental queries, fuzzy queries, spatio-temporal queries, feature storage and organization, are also identiied. In our opinion these are vital issues for the success and development of a multimedia query language.
Object oriented computing is in uencing many areas of computer science including software en gine... more Object oriented computing is in uencing many areas of computer science including software en gineering user interfaces operating systems programming languages and database systems The appeal of object orientation is attributed to its higher levels of abstraction for modeling real world concepts its support for incremental development and its potential for interoperability Despite many advances object oriented computing is still in its infancy and a universally acceptable def inition of an object oriented data model is virtually nonexistent although some standardization e orts are underway This report presents the TIGUKAT object model de nition that is the result of an investigation of object oriented modeling features which are common among earlier proposals along with some distinctive qualities that extend the power and expressibility of this model beyond others The literature recognizes two perspectives of an object model the structural view and the behavioral view Most object ori...
Games are used to evaluate and advance Multiagent and Artificial Intelligence techniques. Most of... more Games are used to evaluate and advance Multiagent and Artificial Intelligence techniques. Most of these games are deterministic with perfect information (e.g. Chess and Checkers). A deterministic game has no chance element and in a perfect information game, all information is visible to all players. However, many real-world scenarios with competing agents are stochastic (non-deterministic) with imperfect information. For two-player zero-sum perfect recall games, a recent technique called Counterfactual Regret Minimization (CFR) computes strategies that are provably convergent to an e-Nash equilibrium. A Nash equilibrium strategy is useful in two-player games since it maximizes its utility against a worst-case opponent. However, for multiplayer (three or more player) games, we lose all theoretical guarantees for CFR. However, we believe that CFR-generated agents may perform well in multiplayer games. To test this hypothesis, we used this technique to create several 3-player limit Tex...
Naive Bayes classifiers, a popular tool for predicting the labels of query instances, are typical... more Naive Bayes classifiers, a popular tool for predicting the labels of query instances, are typically learned from a training set. However, since many training sets contain noisy data, a classifier user may be reluctant to blindly trust a predicted label. We present a novel graphical explanation facility for Naive Bayes classifiers that serves three purposes. First, it transparently explains the reasoning used by the classifier to foster user confidence in the prediction. Second, it enhances the user's understanding of the complex relationships between the features and the labels. Third, it can help the user to identify suspicious training data. We demonstrate these ideas in the context of our implemented web-based system, which uses examples from molecular biology.
Workstation environments have been in use for more than a decade now. Although a network of works... more Workstation environments have been in use for more than a decade now. Although a network of workstations together represents a large amount of aggregate computing power, single users often cannot utilize these resources for their applications. E terprise is a programming environment for designing, coding, debugging, testing, monitoring, profiling and executing programs in a distributed hardware environment. Enterprise code looks like familiar sequential code; the parallelism is expressed graphically. The system automatically inserts the code necessary to handle communication, synchronization and fault tolerance, allowing the rapid construction of correct distributed programs. Enterprise programs run on a network of computers, absorbing the idle cycles on machines. The system supports load balancing, limited process migration, and dynamic distribution of work in environments with changing resource utilization. E terprise offers a costeffective method for increasing the productivity o...
In this paper we consider various spatial relationships that are of general interest for retrievi... more In this paper we consider various spatial relationships that are of general interest for retrieving data from multimedia databases. We present a uni ed representation of spatial objects for both topological and directional relations. Such a representation is based on Allen's temporal interval algebra. We also present a set of spatial inference rules, which allow us to make heterogeneous spatial relation deductions from existing directional and topological relations. For example, if we know A north of B, B overlap with C, and C north of D, then we deriveA above D. Since all the rules are propositional Horn clauses, they can be easily integrated into any multimedia database by using either a simple inference engine or a lookup table.
The Enterprise programming environment supports the development of applications that run concurre... more The Enterprise programming environment supports the development of applications that run concurrently on a network of workstations. This paper describes the object-oriented components of Enterprise, implemented in Smalltalk-80, and their seamless integration with the procedural components, implemented in C. The object-oriented user-interface supports a new anthropomorphic model for parallel computation that eliminates much of the perceived complexity of parallel programs. The object-oriented animation component implements a new animation architecture that supports synchronous and asynchronous events. This allows a user to view the dynamic interactions of the parallel components of a distributed application to simplify performance monitoring and debugging. The Enterprise experience highlights the strengths of object-oriented methodologies both for expressing user models and for implementing related components.
In large extensive form games with imperfect information, Counterfactual Regret Minimization (CFR... more In large extensive form games with imperfect information, Counterfactual Regret Minimization (CFR) is a popular, iterative algorithm for computing approximate Nash equilibria. While the base algorithm performs a full tree traversal on each iteration, Monte Carlo CFR (MCCFR) reduces the per iteration time cost by traversing just a sampled portion of the tree. On the other hand, MCCFR's sampled values introduce variance, and the effects of this variance were previously unknown. In this paper, we generalize MCCFR by considering any generic estimator of the sought values. We show that any choice of an estimator can be used to probabilistically minimize regret, provided the estimator is bounded and unbiased. In addition, we relate the variance of the estimator to the convergence rate of an algorithm that calculates regret directly from the estimator. We demonstrate the application of our analysis by defining a new bounded, unbiased estimator with empirically lower variance than MCCFR...
Modeling moving objects has become a topic of increasing interest in the area of video databases.... more Modeling moving objects has become a topic of increasing interest in the area of video databases. Two key aspects of such modeling are object spatial and temporal relationships. In this paper we introduce an innovative way to represent the trajectory of single moving object and the relative spatio-temporal relations between multiple moving objects. The representation supports a rich set of spatial topological and directional relations. It also supports both quantitative and qualitative user queries about moving objects. Algorithms for matching trajectories and spatio-temporal relations of moving objects are designed to facilitate query processing. These algorithms can handle both exact and similarity matches. We also discuss the integration of our moving object model, based on a video model, in an object-oriented system. Some query examples are provided to further validate the expressiveness of our model.
A "first" implementation of the Modular Smalltalk object-oriented programming language ... more A "first" implementation of the Modular Smalltalk object-oriented programming language is presented. The implementation includes an object-oriented parser, object-oriented representation for code fragments and an object-oriented C-code generator, all implemented in Smalltalk-80. This implementation validates two of the five design principles of the Modular Smalltalk language and provides a vehicle for validating the other three design principles. The macro-based C-code generator is easily adaptable to generating production code in other languages like assembler. In addition, the generation technique applies to source languages other than Modular Smalltalk. The implementation includes an efficient method dispatch based on new extensions to incremental cache table coloring.
Uploads
Papers by Duane Szafron