A method for tuning parameters under a fixed time constraint for a general binocular stereo-vision algorithm is presented. TEACHER 4.2, a generate-and-test system that systematically generates new parameter values by analyzing the results of previous tests and performs limited and controlled tests on the candidates generated using high-speed computers, is discussed. The system is modeled as a statistical selection problem operating under a given time constraint. Results show that the system can find new parameter-value sets which in some cases are better than the ones originally found by extensive hand-tuning and commonly used heuristics, and that different parameter values may be required under different objectives and performance constraints.<>
{"title":"Automated parameter tuning in stereo vision under time constraints","authors":"Steven R. Schwartz, B. Wah","doi":"10.1109/TAI.1992.246359","DOIUrl":"https://doi.org/10.1109/TAI.1992.246359","url":null,"abstract":"A method for tuning parameters under a fixed time constraint for a general binocular stereo-vision algorithm is presented. TEACHER 4.2, a generate-and-test system that systematically generates new parameter values by analyzing the results of previous tests and performs limited and controlled tests on the candidates generated using high-speed computers, is discussed. The system is modeled as a statistical selection problem operating under a given time constraint. Results show that the system can find new parameter-value sets which in some cases are better than the ones originally found by extensive hand-tuning and commonly used heuristics, and that different parameter values may be required under different objectives and performance constraints.<<ETX>>","PeriodicalId":265283,"journal":{"name":"Proceedings Fourth International Conference on Tools with Artificial Intelligence TAI '92","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134537564","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The problem of planning motions offline for a robot in the presence of obstacles via an algorithm that does not proceed by a discretization of the search space involves defining good heuristics to characterize the proximal relationships between two rigid bodies and the development of algorithms for their computation. Several measures are presented, and schemes for their computation are developed. The performances of the algorithms are determined in terms of their running time. The advantage of the fuzzy model is that it lends itself easily to numerical computation involving manipulation of the coordinates of the extreme points.<>
{"title":"Fuzzy distance functions for motion planning","authors":"K. Sridharan, H. Stephanou","doi":"10.1109/TAI.1992.246356","DOIUrl":"https://doi.org/10.1109/TAI.1992.246356","url":null,"abstract":"The problem of planning motions offline for a robot in the presence of obstacles via an algorithm that does not proceed by a discretization of the search space involves defining good heuristics to characterize the proximal relationships between two rigid bodies and the development of algorithms for their computation. Several measures are presented, and schemes for their computation are developed. The performances of the algorithms are determined in terms of their running time. The advantage of the fuzzy model is that it lends itself easily to numerical computation involving manipulation of the coordinates of the extreme points.<<ETX>>","PeriodicalId":265283,"journal":{"name":"Proceedings Fourth International Conference on Tools with Artificial Intelligence TAI '92","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125572833","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Multiple inheritance systems with exceptions in object-oriented languages are examined. Two types of exceptions may be identified: exceptions by cancellation of an inheritance link and exceptions by cancellation of property. For each type of exception, contradictions appear when there are simultaneously several paths of the inheritance graph, which allow inheritance from an object or from a property to take place or not. If certain contradictions are commonly solved by masking, there are contradictions for which no common method of resolution exists. A method for the management of the contradictions in object languages is proposed. This method is based on a computation of the complexity of inheritance paths which produce a contradiction.<>
{"title":"A method for the management of exceptions in multiple inheritance systems","authors":"M. Oussalah, M. Magnan, L. Torrès","doi":"10.1109/TAI.1992.246406","DOIUrl":"https://doi.org/10.1109/TAI.1992.246406","url":null,"abstract":"Multiple inheritance systems with exceptions in object-oriented languages are examined. Two types of exceptions may be identified: exceptions by cancellation of an inheritance link and exceptions by cancellation of property. For each type of exception, contradictions appear when there are simultaneously several paths of the inheritance graph, which allow inheritance from an object or from a property to take place or not. If certain contradictions are commonly solved by masking, there are contradictions for which no common method of resolution exists. A method for the management of the contradictions in object languages is proposed. This method is based on a computation of the complexity of inheritance paths which produce a contradiction.<<ETX>>","PeriodicalId":265283,"journal":{"name":"Proceedings Fourth International Conference on Tools with Artificial Intelligence TAI '92","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129212805","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A nonconventional text character extraction technique, called horizontal-vertical projection (HVP), has been used for feature extraction of numerous alphanumeric symbols. In an earlier paper a classification algorithm based on a 1-D Fourier transform (FT) was used with NVP. An alternative 2-D FT feature extraction technique is presented and shown to improve on the 1-D approach.<>
{"title":"Improving a knowledge based approach for recognition of typed text characters using a 2-D FT","authors":"A. T. Gumahad, N. Bourbakis","doi":"10.1109/TAI.1992.246411","DOIUrl":"https://doi.org/10.1109/TAI.1992.246411","url":null,"abstract":"A nonconventional text character extraction technique, called horizontal-vertical projection (HVP), has been used for feature extraction of numerous alphanumeric symbols. In an earlier paper a classification algorithm based on a 1-D Fourier transform (FT) was used with NVP. An alternative 2-D FT feature extraction technique is presented and shown to improve on the 1-D approach.<<ETX>>","PeriodicalId":265283,"journal":{"name":"Proceedings Fourth International Conference on Tools with Artificial Intelligence TAI '92","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126791200","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A method of incorporating real-time semantics in production rules, thus making them suitable for representing knowledge of a real-time expert system, is presented. It is intended that the knowledge representation should allow the incorporation of time-dependent heuristics and dynamic models in the knowledge base. An illustration of this method, the NetManager expert system, is noted.<>
{"title":"Knowledge with real-time semantics","authors":"Annie Z. Shamsudin, T. Dillon","doi":"10.1109/TAI.1992.246453","DOIUrl":"https://doi.org/10.1109/TAI.1992.246453","url":null,"abstract":"A method of incorporating real-time semantics in production rules, thus making them suitable for representing knowledge of a real-time expert system, is presented. It is intended that the knowledge representation should allow the incorporation of time-dependent heuristics and dynamic models in the knowledge base. An illustration of this method, the NetManager expert system, is noted.<<ETX>>","PeriodicalId":265283,"journal":{"name":"Proceedings Fourth International Conference on Tools with Artificial Intelligence TAI '92","volume":"141 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124224379","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The neural network described performs classification of software metrics. It is a three-layer, error back-propagation network. Using historical data, the neural network learns the relationship between certain metrics and a particular classification. The neural network selects the classification which best fits the input metrics. The capability of neural networks to classify nonlinearly separable problem spaces gives them an advantage over tree-based and linear network-based classification methods. When applied to actual software metrics, the neural network correctly classified 100% of the data presented.<>
{"title":"Metric-based neural network classification tool for analyzing large-scale software","authors":"R. Paul","doi":"10.1109/TAI.1992.246366","DOIUrl":"https://doi.org/10.1109/TAI.1992.246366","url":null,"abstract":"The neural network described performs classification of software metrics. It is a three-layer, error back-propagation network. Using historical data, the neural network learns the relationship between certain metrics and a particular classification. The neural network selects the classification which best fits the input metrics. The capability of neural networks to classify nonlinearly separable problem spaces gives them an advantage over tree-based and linear network-based classification methods. When applied to actual software metrics, the neural network correctly classified 100% of the data presented.<<ETX>>","PeriodicalId":265283,"journal":{"name":"Proceedings Fourth International Conference on Tools with Artificial Intelligence TAI '92","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129135993","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
As part of developing the CLIPS-based debugging tool MIRO, and because there has been relatively little exploration of the debugging process for such programs, the problems of debugging data-driven rule-based programs in general are defined and analyzed. How explanation can be used as a tool to ease the debugging process is discussed, and how it can suggest causes of faults in answer to questions asking why a particular fault occurred is examined. An algorithm that provides part of the explanation for a rule's failure to fire is discussed.<>
{"title":"Suggesting causes of faults in data-driven rule-based systems","authors":"Sharon M. Tuttle, C. Eick","doi":"10.1109/TAI.1992.246438","DOIUrl":"https://doi.org/10.1109/TAI.1992.246438","url":null,"abstract":"As part of developing the CLIPS-based debugging tool MIRO, and because there has been relatively little exploration of the debugging process for such programs, the problems of debugging data-driven rule-based programs in general are defined and analyzed. How explanation can be used as a tool to ease the debugging process is discussed, and how it can suggest causes of faults in answer to questions asking why a particular fault occurred is examined. An algorithm that provides part of the explanation for a rule's failure to fire is discussed.<<ETX>>","PeriodicalId":265283,"journal":{"name":"Proceedings Fourth International Conference on Tools with Artificial Intelligence TAI '92","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130284454","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
An X window system interface, the Neural Shell Version 3.5, which simplifies the use of neural-network simulators for neural-network research is presented. This interface can be customized by the user as additional simulators and toolkit programs are developed and added to the system. The Neural Shell runs on workstations and additionally supports the execution of large neural-network simulations on remote high-performance computers. Typically, users develop prototype neural networks on local workstations and then perform more extensive simulations on remote machines. The functional capabilities of the configurable user interface are described.<>
给出了一个X窗口系统接口Neural Shell Version 3.5，它简化了神经网络仿真器在神经网络研究中的使用。当开发并添加到系统中的附加模拟器和工具包程序时，用户可以自定义该接口。Neural Shell可以在工作站上运行，还支持在远程高性能计算机上执行大型神经网络模拟。通常，用户在本地工作站开发原型神经网络，然后在远程机器上进行更广泛的模拟。介绍了可配置用户界面的功能。
{"title":"A user customizable X window system interface for simulation of neural networks","authors":"T. E. Little, S. Ahalt","doi":"10.1109/TAI.1992.246449","DOIUrl":"https://doi.org/10.1109/TAI.1992.246449","url":null,"abstract":"An X window system interface, the Neural Shell Version 3.5, which simplifies the use of neural-network simulators for neural-network research is presented. This interface can be customized by the user as additional simulators and toolkit programs are developed and added to the system. The Neural Shell runs on workstations and additionally supports the execution of large neural-network simulations on remote high-performance computers. Typically, users develop prototype neural networks on local workstations and then perform more extensive simulations on remote machines. The functional capabilities of the configurable user interface are described.<<ETX>>","PeriodicalId":265283,"journal":{"name":"Proceedings Fourth International Conference on Tools with Artificial Intelligence TAI '92","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121388966","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The idea of competitive learning for pattern-recognition applications is introduced. A brief review of two competitive learning models, T. Kohonen's self-organizing feature maps (1982, 1989) and S. Grossberg's ART networks (1987), is presented. Neural-net-based partitioning algorithms for learning control are introduced. A simulation study, of these algorithms incorporated into the BOXES machine learning control system is reported. Simulation results are presented and performance comparisons are made, using the BOXES algorithm as the standard, with the new neural-net-based partitioning method. The original BOXES partitioning method of fixed threshold quantization of state-space variables was used in the BOXES algorithm learning trials.<>
{"title":"Neural network based competitive learning for control","authors":"B. Zhang, E. Grant","doi":"10.1109/TAI.1992.246409","DOIUrl":"https://doi.org/10.1109/TAI.1992.246409","url":null,"abstract":"The idea of competitive learning for pattern-recognition applications is introduced. A brief review of two competitive learning models, T. Kohonen's self-organizing feature maps (1982, 1989) and S. Grossberg's ART networks (1987), is presented. Neural-net-based partitioning algorithms for learning control are introduced. A simulation study, of these algorithms incorporated into the BOXES machine learning control system is reported. Simulation results are presented and performance comparisons are made, using the BOXES algorithm as the standard, with the new neural-net-based partitioning method. The original BOXES partitioning method of fixed threshold quantization of state-space variables was used in the BOXES algorithm learning trials.<<ETX>>","PeriodicalId":265283,"journal":{"name":"Proceedings Fourth International Conference on Tools with Artificial Intelligence TAI '92","volume":"237 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116432226","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Various roles that high-level or abstract models of expertise could play in improving the process of knowledge acquisition are reviewed. A detailed analysis of the knowledge acquisition process is presented. This analysis allows the different aspects of the process where using problem-solving models might be potentially beneficial to be identified. A synthetic view of these potential roles is given. An attempt to make these roles operational with a knowledge engineering workbench, limitations of the approach, and related work are discussed.<>
{"title":"Various uses of problem solving models for knowledge acquisition","authors":"F. Ramparany","doi":"10.1109/TAI.1992.246434","DOIUrl":"https://doi.org/10.1109/TAI.1992.246434","url":null,"abstract":"Various roles that high-level or abstract models of expertise could play in improving the process of knowledge acquisition are reviewed. A detailed analysis of the knowledge acquisition process is presented. This analysis allows the different aspects of the process where using problem-solving models might be potentially beneficial to be identified. A synthetic view of these potential roles is given. An attempt to make these roles operational with a knowledge engineering workbench, limitations of the approach, and related work are discussed.<<ETX>>","PeriodicalId":265283,"journal":{"name":"Proceedings Fourth International Conference on Tools with Artificial Intelligence TAI '92","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126520154","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：