Proposes efficient postprocessing algorithms for error correction in handwritten Hangul (Korean script) address and human name recognition. As the load on the character recognizer for the recognition of the administra...
详细信息
Proposes efficient postprocessing algorithms for error correction in handwritten Hangul (Korean script) address and human name recognition. As the load on the character recognizer for the recognition of the administrative district part in addresses was reduced by restricting the candidate characters to be matched based on a hierarchical address lexicon, the processing speed and recognition rate were greatly improved. Also, the misrecognition results from the character recognizer were corrected by using efficient postprocessing algorithms based on backtracking. For the recognition of the human name part, misrecognition of human names could be effectively corrected by combining the a priori probability and the confusion probability of each character making up the human names.< >
A knowledge-based learning system is developed to demonstrate that intelligently selecting a subset of examples based on domain knowledge for rule induction can be more productive than using all provided examples. Kno...
详细信息
作者：
KING, JFBARTON, DEJ. Fred King:is the manager of the Advanced Technology Department for Unisys in Reston
Virginia. He earned his Ph.D. in mathematics from the University of Houston in 1977. He has been principal investigator of research projects in knowledge engineering pattern recognition and heuristic problem-solving. Efforts include the development of a multi-temporal multispectral classifier for identifying graincrops using LANDSAT satellite imagery data for NASA. Also as a member of the research team for a NCI study with Baylor College of Medicine and NASA he helped develop techniques for detection of carcinoma using multispectral microphotometer scans of lung tissue. He established and became technical director of the AI Laboratory for Ford Aerospace where he developed expert scheduling modeling and knowledge acquisition systems for NASA. Since joining Unisys in 1985 he has led the development of object-oriented programming environments blackboard architectures data fusion techniques using neural networks and intelligent data base systems. Douglas E. Barton:is manager of Logistics Information Systems for Unisys in Reston
Virginia. He earned his B.A. degree in computer science from the College of William and Mary in 1978 and did postgraduate work in London as a Drapers Company scholar. Since joining Unisys in 1981 his work has concentrated on program management and software engineering of large scale data base management systems and design and implementation of knowledge-based systems in planning and logistics. As chairman of the Logistics Data Subcommittee of the National Security Industrial Association (NSIA) he led an industry initiative which examined concepts in knowledge-based systems in military logistics. His responsibilities also include evaluation development and tailoring of software engineering standards and procedures for data base and knowledge-based systems. He is currently program manager of the Navigation Information Management System which provides support to the Fleet Ballistic Missile Progr
A valuable technique during concept development is rapid prototyping of software for key design components. This approach is particularly useful when the optimum design approach is not readily apparent or several know...
详细信息
A valuable technique during concept development is rapid prototyping of software for key design components. This approach is particularly useful when the optimum design approach is not readily apparent or several known alternatives need to be rapidly evaluated. A problem inherent in rapid prototyping is the lack of a "target system" with which to interface. Some alternatives are to develop test driver libraries, integrate the prototype with an existing working simulator, or build one for the specific problem. This paper presents a unique approach to concept development using rapid prototyping for concept development and scenario-based simulation for concept verification. The rapid prototyping environment, derived from artificial intelligence technology, is based on a blackboard architecture. The rapid prototype simulation capability is provided through an object-oriented modeling environment. It is shown how both simulation and blackboard technologies are used collectively to rapidly gain insight into a tenacious problem. A specific example will be discussed where this approach was used to evolve the logic of a mission controller for an autonomous underwater vehicle.
作者：
Peter BockThe University of Ulm
Research Institute for Applied Knowledge processing (FAW) Ulm West Germany and Department of Electrical Engineering and Computer Science The George Washington University Washington DC
An alternative to preprogrammed rule-based Artificial Intelligence is a hierarchical network of cellular automata which acquire their knowledge through learning based on a series of trial-and-error interactions with a...
ISBN:
(纸本)9780897913485
An alternative to preprogrammed rule-based Artificial Intelligence is a hierarchical network of cellular automata which acquire their knowledge through learning based on a series of trial-and-error interactions with an evaluating Environment, much as humans do. The input to the hierarchical network is provided by a set of sensors which perceive the external world. Based upon this perceived information and past experience (memory), the learning automata synthesize collections of trial responses. Periodically the automata estimate the effectiveness of these collections using either internal evaluations (unsupervised learning) or external evaluations from the Environment (supervised learning), modifying their memories accordingly. Known as Collective Learning Systems Theory, this paradigm has been applied to many sophisticated gaming problems, demonstrating robust learning and dynamic *** on a versatile architecture for massively parallel networks of processors for Collective Learning Systems, a Transputer-based parallel-processing image processing engine comprising 32 learning cells and 32 non-learning cells has been applied to a sophisticated image processing task: the scale-invariant and translation-invariant detection of anomalous features in otherwise “normal” images. In cooperation with Robert Bosch GmbH, this engine is currently being constructed and tested under the direction of the author at the Research Institute for Applied knowledge Processing (FAW-Ulm) as Project ALIAS: Adaptive Learning Image Analysis System. Initial results indicate excellent detection, discrimination, and localization of anomalies.
Learning from inconclusive data is an important problem that has not been addressed in the concept learning literature. In this paper, we define inconclusiveness and illustrate why ID3-like algorithms are bound to res...
详细信息
This paper describes a modularized AI system being built to help improve electromagnetic compatibility (EMC) among shipboard topside equipment and their associated systems. CLEER is intended to act as an easy to use i...
详细信息
This paper describes a modularized AI system being built to help improve electromagnetic compatibility (EMC) among shipboard topside equipment and their associated systems. CLEER is intended to act as an easy to use integrator of existing expert knowledge and pre-existing data bases and large scale analytical models. Due to these interfaces; to the need for portability of the software; and to artificial intelligence related design requirements (such as the need for spatial reasoning, expert data base management, model base management, track-based reasoning, and analogical (similar ship) reasoning) it was realized that traditional expert system shells would be inappropriate, although relatively off-the-shelf AI technology could be incorporated. In the same vein, the rapid prototyping approach to expert system design and knowledgeengineering was not pursued in favor of a rigorous systems engineering methodology. The critical design decisions affecting CLEER's development are summarized in this paper along with lessons learned to date all in terms of “how,” “why,” and “when” specific features are being developed.
This paper describes the meta-level control system of a program (Dominic) for parametric design of mechanical components by iterative redesign. We view parametric design as search, and thus Dominic is a hill climbing ...
详细信息
暂无评论