conference on Analysis of Neural Network Applications最新文献

{"title":"Defect prediction with neural networks","authors":"R. Stites, Bryan Ward, Robert Henry Walters","doi":"10.1145/106965.106970","DOIUrl":"https://doi.org/10.1145/106965.106970","url":null,"abstract":"The industrial and scientific world abound with problems that are poorly un&rstood or for which apparent anomalous conditions exist. Artificial Neural Networks are utilized with conventional techniques to extract salient features and relationships which are non-linear in nature. Defect causality in a large continuous flow chemical process is investigated. Significant gains in the prediction of defects over traditional statistical methods are achieved.","PeriodicalId":359315,"journal":{"name":"conference on Analysis of Neural Network Applications","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115205931","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}

{"title":"A discrete-time neural network multitarget tracking data association algorithm","authors":"O. Olurotimi","doi":"10.1145/106965.105255","DOIUrl":"https://doi.org/10.1145/106965.105255","url":null,"abstract":"This paper describes an alternative solution to that proposed earlier by Sengupta and Iltis (1989). The earlier work exploited widely known stability results for analog continuous-time neural networks. Such results are not known in general for analog, discrete-time networks. Therefore it is not a straightforward issue to transport a continuous-time solution into a purely discrete-time domain. In this paper, we define a particular analog, discrete-time network which is structurally similar to the earlier continuous-time form. We show that, with a weight structure analogous to that earlier obtained from Liapunov function considerations, and with a few mild constraints, this discrete-time network has qualitative stability properties that can be similarly exploited. Permission to copy without fee all or part of this material is granted provided that tbe copies are not made or distributed for direct commercial advantage, the ACM copyright notice and the title of the publication and its date appear, and notice is given that copying is by permission of the Association for Computing Machinery. To copy otherwise, or to republish requires a fee and/or specific permission. The resulting system is applied, in an analogous way to that of Sengupta and Iltis, to the data association problem in multitarget tracking. An advantage of the proposed approach is that it is more amenable to inclusion in clocked, or digital systems. Such an implementation will also run much faster than a discretized version of the earlier algorithm.","PeriodicalId":359315,"journal":{"name":"conference on Analysis of Neural Network Applications","volume":"89 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131802597","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}

{"title":"Optimization in cascaded Boltzmann machines with a temperature gradient: an alternative to simulated annealing","authors":"J. Coughlin, R. Baran","doi":"10.1145/106965.106968","DOIUrl":"https://doi.org/10.1145/106965.106968","url":null,"abstract":"Boltzmann machines can be series-coupled with one-way retinotopic connections to produce good (suboptimal) solutions to classic optimization problems. Each isothermal Bokzmann machine in the series has a lower temperature than the one which drives it through excitatory spanning links. This scheme is an alternative to the usual simulated annealing approach in which a single network is cooled over the course of time. The simplest case features two identical subnets, one at a positive temperature and the other at zero temperature. Its performance in solving n-to-n assignment probems is compared to that obtained by simulated annealing with a geometric cooling schedule.","PeriodicalId":359315,"journal":{"name":"conference on Analysis of Neural Network Applications","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126474867","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}

{"title":"Spacial classification and multi-spectral fusion with neural networks","authors":"C. Harston","doi":"10.1145/106965.105257","DOIUrl":"https://doi.org/10.1145/106965.105257","url":null,"abstract":"Neural networks classified a thematic mapper LandSat 4 multi-spectral image of the area surrounding Murfree=boroS Tennessee. Back propagation neural network= were trained to identify different land types. Six area% were partially classified with individual networks for each band. The results were combined/fused with a network to categorize all six areas. In another analysis~ contiguous points of Maximum– likelihood classifications (MLC) were reclassified by a neural network. Amoung other things~ this network learned to distinguish between buildings and rocks that were classified the same by the MLC. Clearly Permissiontocopy withoutfeeallorpart ofthismatenalis granted provided that the copies are not made or distributed for direct commercial advantage, the ACM copyright notice and tie title of dre publication and its date appear, and notice is given that copying is by permissionoftheAssoeiation forComputingMachinery. Tocopy otherwise, or to republish requires a fee and/or specific permission. neural networks can be used for multi-spectral classification. The combination of MLC and neural techniques is productive. Real time multi–spectral processing may be possible with neural network hardware. INTRCHXJCTION Multi-spectral remotely sensed data i% used to classify areas of the earth. Urban areas can be identified~ crops quantified, forests evaluated and oceans studied. Multi– spectral classification techniques are well proven and commercially viable (Richa%an~","PeriodicalId":359315,"journal":{"name":"conference on Analysis of Neural Network Applications","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125286118","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}

{"title":"On the implementation of RBF technique in neural networks","authors":"M. Musavi, K. B. Faris, Khue Hiang Chan, W. Ahmed","doi":"10.1145/106965.105254","DOIUrl":"https://doi.org/10.1145/106965.105254","url":null,"abstract":"An approach for the implementation of the Radial Basis Function (RBF) technique is presented and applied to a net work of the appropriate architecture. The paper explores the extent to which the number of RBF nodes can be reduced without significantly affecting the overall training error. This is accomplished through an effective clustering algorithm that shall be described in detail. Emphasis is also placed on the problems faced by a technique that has been proved superior to the more traditional training algorithms, particularly in terms of processing speed and solvability of nonlinear patterns. Solutions are consequently proposed in view of making RBF a more efficient method for interpolation and classification purposes.","PeriodicalId":359315,"journal":{"name":"conference on Analysis of Neural Network Applications","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125207407","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}

{"title":"Protein classification using a neural network database system","authors":"Cathy H. Wu, T. Chang","doi":"10.1145/106965.105260","DOIUrl":"https://doi.org/10.1145/106965.105260","url":null,"abstract":"Neural networks are being applied to a widely expanding area of applications, including the biological applications of protein structure prediction and DNA sequence analysis. This paper describes a novel application of neural networks to the classification of the immense amounts of sequencing data being generated by the Human Genome Project and genetic engineering research. The protein classification is an alternative approach to the large database search problem so that the search time is not constrained by the database size. Previously, we have implemented a prototype protein classification system, PRO CANS, and demonstrated rapid and accurate allocations of 30 protein classes. This research scales up the pilot system into a “neurat database” system and aims at the classification of unknown protein sequences into 2,350 protein superfamilies (classes) currently being identified in the PIR (Protein Identification Resources) protein sequence database. The neural network protein database (NNPDB) system involves two major design principles: (a) a sequence encoding schema to effectively retrieve salient information from sequence strings, and (b) a modular network architecture to store the huge amount of training patterns. The complete NNPDB program, which includes preprocessor for sequence encoding, neural network for classification, and postprocessor for report generation, has been implemented on a CONVEX/CRAY computer platform. The NNPDB system is developed incrementally by training and optimizing each network module. After the training of 200 to 13,000 CRAY CPU seconds for the network modules, the system is able to predict within three CPU F’mnissiott to copy without fee all or part of this material is granted provided that Ute cople! are nol made or disrribmed for direct commercial advantage, the ACM copyright notice and the title of the publication and its date appear, and notice is given that co ying is by $ permission of rhe Association for Computing Machinery. o copy otfrerwise, or to repubtish requires a fee end/or specific permission. seconds with a 90 to 99% accuracy for the two protein groups tested, the electron transfer proteins and the oxidoreductases. In addition to the accuracy and speed of classification, the system architecture permits the identification of salient sequence information and flexible database growth and update. The neural database, which consists of a set of weight matrices of the networks, can be portable to other computers for speedy on-line anatysis of new sequences, and directly benefits the biology community. Furthermore, the system design should be easily adaptable for the information processing of other large and complex domains.","PeriodicalId":359315,"journal":{"name":"conference on Analysis of Neural Network Applications","volume":"407 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116526694","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}

{"title":"A neural network for target classification using passive sonar","authors":"R. Baran, J. Coughlin","doi":"10.1145/106965.106969","DOIUrl":"https://doi.org/10.1145/106965.106969","url":null,"abstract":"This concerns the design, tratilng, test and evaluation of a feed-forward neural network for classifying acoustic signals emitted by ships in transit by an omnidirectional hydrophore. Relatively noisy surface ships, moving rapidly at medium to long range, emit signals which superficially resemble those of quieter submarines, moving more slowly and closer to the listening device. The neural network approach is motivated by an obvious analogy to the sonar classifier of German and Sejnowski, who trained a neural network to classify active sonar returns from two undersea objects. The present problem can be solved by a similar network architecture, the outputs indicating which target type (if any) is present. The inputs represent the evolution of spectral densities for each of a number of time lags. Yet the number of target types and encounter geometries is far greater than could possibly be covered in any representative way by a training set comprised of real world data, Thus the task is to connect the network to a high fidelity, model-based digital simulator and to show that, by training on the output of the simulator, the neural network can learn to pass realistic tests. Permission to copy without fee all or part of this material is gmrtted provided that the copies are not made or distributed for direct commercial advantage, the ACM copyright notice and the title of the pubtieation and its date appear, and notice is given that copying is by permission of the Association for Computing Mach~:ry. TOCOPY otherwise,or to republish requires a fee andlor spectilc perrms sion. This describes a neural network design-and-testing exercise based on a simplistic model that captures a few of the salient features of the problem.","PeriodicalId":359315,"journal":{"name":"conference on Analysis of Neural Network Applications","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122000723","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}

{"title":"Pattern recognition with a pulsed neural network","authors":"J. Dayhoff","doi":"10.1145/106965.106966","DOIUrl":"https://doi.org/10.1145/106965.106966","url":null,"abstract":"This paper describes pulse transmission (PT) neural networks and provides a series of analytical and experimental findings regarding their abilities to perform pattern-mapping and temporal integration. We show the synaptic equations and the frequency transfer characteristics of the synapses. A network is described that performs the Xor function via pulses and that amplifies its output rate by means of synchronous groups assemblies of units that fire or burst simult aneously. Another network is described that does coarse-grid pattern recognition with pulses. This network performs temporal integration of pattern fragments over a period of time, and recognizes patterns that are obscured by noisy but changing foreground objects. Similarity is shown between certain pulsed networks in frequency operating modes and particular configurations of recurrent networks. A general class of pulsed neural networks is described.","PeriodicalId":359315,"journal":{"name":"conference on Analysis of Neural Network Applications","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130299961","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}

{"title":"Neural network-based decision support for incomplete database systems: Knowledge acquisition and performance analysis","authors":"Bo Jin, A. Hurson, L. Miller","doi":"10.1145/106965.105258","DOIUrl":"https://doi.org/10.1145/106965.105258","url":null,"abstract":"This paper investigates the design and implementation of a knowledge acquisition module as a part of a decision support system for handling large incomplete databases. Maybe algebra and attribute maybe algebra operations were introduced and implemented as extensions to the relational algebra operations. These operations give a user the opportunity to investigate the set of data containing null values (i.e. incomplete tuples) to draw his/her own conclusions. However, some of these operations may generate enormous and/or erroneous data. Moreover, under the maybe operations, some of the resultant data may not provide useful information for the user. A decision support system based on an artificial neural network is proposed to increase data qufllty in the presence of missing/incomplete information. Based on the learned knowledge, the neural network can filter out the undesirable data. In the proposed decision support system, a knowledge acquisition module plays a vital role in generating training data (training pairs). Data semantics of the underlying databases (e.g. data dependency conditions) is the main source from which such a knowledge can be acquired. It is demonstrated that the knowledge acquisition can be accomplished by using a two-level hierarchical structured model. Finally, some simulation results are presented to demonstrate the feasibility and performance of the proposed knowledge acquisition medule. Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the ACM copyright notice and the titte of the publication and its date appear, and notice is given that eo ying is by ! permission of the Association for Computing Machinery. ocopy otherwise, or to republish requires a fee and/or specific pemussion. ** Dep~ent of Computer Science Iowa State University Ames, IA 50011","PeriodicalId":359315,"journal":{"name":"conference on Analysis of Neural Network Applications","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131048285","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}

{"title":"Neural network process control","authors":"M. Piovoso, A. J. Owens","doi":"10.1145/106965.105256","DOIUrl":"https://doi.org/10.1145/106965.105256","url":null,"abstract":"Neural Networks are increasingly finding engineering applications. Most early applications were in the areas of pattern recognition and modeling. This paper shows how neural network models can be used in process control. Two separate techniques are illustrated, each with a specific example application. One involves using the network itself as the inverse model, by fixing the neural network weights and training on the inputs to give the desired output pattern. The other suggests using the pattern recognition ability of a neural network to identify an appropriate lower order linear model to use for controller design.","PeriodicalId":359315,"journal":{"name":"conference on Analysis of Neural Network Applications","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121607850","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}