Pub Date : 2000-12-01DOI: 10.1109/TENCON.2000.893529
M. Fukumi, Y. Mitsukura, N. Akamatsu
A new rule generation method from neural networks is presented. A neural network (NN) is formed using a genetic algorithm (GA) with virus infection and deterministic mutation to represent regularities in training data. This method utilizes a modular structure in GA. Each module learns a different neural network architecture, such as sigmoid and a higher order neural networks. Those chromosome information is communicated to the other modules by the virus infection. The higher order units are connected to an output unit or hidden units. By using these architectures, rules can be extracted. The results of computer simulations show that this approach can generate obvious network architectures and as a result simple rules.
{"title":"A new genetic approach to universal rule generation from trained neural networks","authors":"M. Fukumi, Y. Mitsukura, N. Akamatsu","doi":"10.1109/TENCON.2000.893529","DOIUrl":"https://doi.org/10.1109/TENCON.2000.893529","url":null,"abstract":"A new rule generation method from neural networks is presented. A neural network (NN) is formed using a genetic algorithm (GA) with virus infection and deterministic mutation to represent regularities in training data. This method utilizes a modular structure in GA. Each module learns a different neural network architecture, such as sigmoid and a higher order neural networks. Those chromosome information is communicated to the other modules by the virus infection. The higher order units are connected to an output unit or hidden units. By using these architectures, rules can be extracted. The results of computer simulations show that this approach can generate obvious network architectures and as a result simple rules.","PeriodicalId":396649,"journal":{"name":"2000 TENCON Proceedings. Intelligent Systems and Technologies for the New Millennium (Cat. No.00CH37119)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2000-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130926947","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}
Pub Date : 2000-11-24DOI: 10.1109/TENCON.2000.893678
N. Noor, O. M. Rijal, O.M. Bader, P. Thon
A requirement for good wire bonding is the existence of intermetallics between the gold ball and bond pad. A digital image of the surface of the gold ball that was attached to the bond pad was obtained and analysed with MATLAB. The 'counting method' (CM) is proposed for determining the percentage of intermetallics formation. Four experiments were performed to investigate the reliability of CM, the effect of light-offset, and the performance of a quality assistant (QA) in his ability to visually estimate (the visual method) the intermetallics coverage. The main results indicate that the visual method is inconsistent and less accurate than the counting method. The QA performs well for high intermetallics coverage, whilst for an approximately 50% intermetallics coverage the QA tends to overestimate. Critical remarks related to modeling the digital image are presented.
{"title":"A vision problem in wire bonding","authors":"N. Noor, O. M. Rijal, O.M. Bader, P. Thon","doi":"10.1109/TENCON.2000.893678","DOIUrl":"https://doi.org/10.1109/TENCON.2000.893678","url":null,"abstract":"A requirement for good wire bonding is the existence of intermetallics between the gold ball and bond pad. A digital image of the surface of the gold ball that was attached to the bond pad was obtained and analysed with MATLAB. The 'counting method' (CM) is proposed for determining the percentage of intermetallics formation. Four experiments were performed to investigate the reliability of CM, the effect of light-offset, and the performance of a quality assistant (QA) in his ability to visually estimate (the visual method) the intermetallics coverage. The main results indicate that the visual method is inconsistent and less accurate than the counting method. The QA performs well for high intermetallics coverage, whilst for an approximately 50% intermetallics coverage the QA tends to overestimate. Critical remarks related to modeling the digital image are presented.","PeriodicalId":396649,"journal":{"name":"2000 TENCON Proceedings. Intelligent Systems and Technologies for the New Millennium (Cat. No.00CH37119)","volume":"203 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2000-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134439602","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}
Pub Date : 2000-11-01DOI: 10.1109/TENCON.2000.892312
C. Loo, M. Rajeswari
In this paper, a particular class of modular network called Growing Multi-Experts network is presented. The core idea of Multi-Experts network shares a conceptual link with divide-and conquer methodology. In this regard, the task of approximating a complicated nonlinear function can be split up among local experts. In other words, the problem space is decomposed into overlapping regions and local experts approximate the data in every region. Growing Multi-Experts network may be subjected to over-fitting problem if the number of experts assigned to the network is not properly defined. The proposed Growing Multi-Experts network is shown to be able to detect redundant experts that exist in the network and suppress the disturbance caused by the redundant experts. A Redundant Experts Removal algorithm is proposed in attempt to remove the redundant experts from the network. This algorithm has been tested with the well-known gas furnace data of Box and Jenkins to compare its performance with other statistical and fuzzy modeling approaches. It is observed that Growing Multi-Experts Network is able to learn the network structure incrementally and reach a good approximation level while maintaining a minimal architecture. The performance of the model compares favorably with these existing techniques.
{"title":"Growing multi-experts network","authors":"C. Loo, M. Rajeswari","doi":"10.1109/TENCON.2000.892312","DOIUrl":"https://doi.org/10.1109/TENCON.2000.892312","url":null,"abstract":"In this paper, a particular class of modular network called Growing Multi-Experts network is presented. The core idea of Multi-Experts network shares a conceptual link with divide-and conquer methodology. In this regard, the task of approximating a complicated nonlinear function can be split up among local experts. In other words, the problem space is decomposed into overlapping regions and local experts approximate the data in every region. Growing Multi-Experts network may be subjected to over-fitting problem if the number of experts assigned to the network is not properly defined. The proposed Growing Multi-Experts network is shown to be able to detect redundant experts that exist in the network and suppress the disturbance caused by the redundant experts. A Redundant Experts Removal algorithm is proposed in attempt to remove the redundant experts from the network. This algorithm has been tested with the well-known gas furnace data of Box and Jenkins to compare its performance with other statistical and fuzzy modeling approaches. It is observed that Growing Multi-Experts Network is able to learn the network structure incrementally and reach a good approximation level while maintaining a minimal architecture. The performance of the model compares favorably with these existing techniques.","PeriodicalId":396649,"journal":{"name":"2000 TENCON Proceedings. Intelligent Systems and Technologies for the New Millennium (Cat. No.00CH37119)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2000-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114420848","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}
Pub Date : 2000-09-27DOI: 10.1109/TENCON.2000.892227
R. Rahman, S. Ikhsan, H. M. Supian, J. Fizik, F. Sains
Optical fibres are well known for their importance as transporter of information. However, much has been said and proved that optical fibres are essential and play their role as producers of information, that is as sensors. Recent research has proved that fibre optic sensors can be integrated in structure such as multistorey buildings, dams and bridges with the capability to monitor their strength and integrity. One component that is increasingly being used in communication and sensing is Bragg grating. A study in the implementation of a fibre optics Bragg grating (FOBG) sensors system is discussed in this paper. This study concentrates on the issues related to the utilization of FOBG sensors in smart structures.
{"title":"Fibre optic Bragg grating sensors: a new technology for smart structure monitoring in Malaysia","authors":"R. Rahman, S. Ikhsan, H. M. Supian, J. Fizik, F. Sains","doi":"10.1109/TENCON.2000.892227","DOIUrl":"https://doi.org/10.1109/TENCON.2000.892227","url":null,"abstract":"Optical fibres are well known for their importance as transporter of information. However, much has been said and proved that optical fibres are essential and play their role as producers of information, that is as sensors. Recent research has proved that fibre optic sensors can be integrated in structure such as multistorey buildings, dams and bridges with the capability to monitor their strength and integrity. One component that is increasingly being used in communication and sensing is Bragg grating. A study in the implementation of a fibre optics Bragg grating (FOBG) sensors system is discussed in this paper. This study concentrates on the issues related to the utilization of FOBG sensors in smart structures.","PeriodicalId":396649,"journal":{"name":"2000 TENCON Proceedings. Intelligent Systems and Technologies for the New Millennium (Cat. No.00CH37119)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2000-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123501763","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}
Pub Date : 2000-09-25DOI: 10.1109/TENCON.2000.893584
Ahmad Idil Abdul Rahman, Sheikh Hussein Shaikh Salleh, A. Sha'ameri, Syed Abdul Rahman AI-Attas
Speech is considered as a time-varying signal since the parameters of the signal such as the amplitude, frequency and phase varies in time. Segmenting a duration of captured speech into analysis frames of 20 msecs ensures the assumption of stationarity. If a captured speech segment representing a word that may last for 600 msec, then a total of 30 analysis frames are required to the word. Due to the possibility that adjacent frames are identical, then it would be of interest to combine these frames into a single long frame. The interval where adjacent frames have identical parameters is referred as the time-invariant interval (TII). It is of interest to determine these intervals and two methods presented are the instantaneous energy and frequency estimation (IEFE) and localized time correlation (LTC) function. A comparison is made in the accuracy in the TII estimate for a set of speech samples.
{"title":"Signal segmentation and its application in the feature extraction of speech","authors":"Ahmad Idil Abdul Rahman, Sheikh Hussein Shaikh Salleh, A. Sha'ameri, Syed Abdul Rahman AI-Attas","doi":"10.1109/TENCON.2000.893584","DOIUrl":"https://doi.org/10.1109/TENCON.2000.893584","url":null,"abstract":"Speech is considered as a time-varying signal since the parameters of the signal such as the amplitude, frequency and phase varies in time. Segmenting a duration of captured speech into analysis frames of 20 msecs ensures the assumption of stationarity. If a captured speech segment representing a word that may last for 600 msec, then a total of 30 analysis frames are required to the word. Due to the possibility that adjacent frames are identical, then it would be of interest to combine these frames into a single long frame. The interval where adjacent frames have identical parameters is referred as the time-invariant interval (TII). It is of interest to determine these intervals and two methods presented are the instantaneous energy and frequency estimation (IEFE) and localized time correlation (LTC) function. A comparison is made in the accuracy in the TII estimate for a set of speech samples.","PeriodicalId":396649,"journal":{"name":"2000 TENCON Proceedings. Intelligent Systems and Technologies for the New Millennium (Cat. No.00CH37119)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2000-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123973365","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}
Pub Date : 2000-09-24DOI: 10.1109/TENCON.2000.892311
N. Ismail, T. C. Fai, M.S. Salif
This paper describes a methodology to integrate design and manufacturing using a feature recognition technique for achieving some concurrent engineering goals. The input of the feature recognition is the boundary representation (B-rep) of solid models. The features considered in this work are protrusion and depression types of prismatic components. The features' faces are planar, cylindrical and hybrid. The recognized features are to be used for a downstream manufacturing application.
{"title":"Form features for concurrent engineering","authors":"N. Ismail, T. C. Fai, M.S. Salif","doi":"10.1109/TENCON.2000.892311","DOIUrl":"https://doi.org/10.1109/TENCON.2000.892311","url":null,"abstract":"This paper describes a methodology to integrate design and manufacturing using a feature recognition technique for achieving some concurrent engineering goals. The input of the feature recognition is the boundary representation (B-rep) of solid models. The features considered in this work are protrusion and depression types of prismatic components. The features' faces are planar, cylindrical and hybrid. The recognized features are to be used for a downstream manufacturing application.","PeriodicalId":396649,"journal":{"name":"2000 TENCON Proceedings. Intelligent Systems and Technologies for the New Millennium (Cat. No.00CH37119)","volume":"172 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2000-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115467782","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}
Pub Date : 2000-09-24DOI: 10.1109/TENCON.2000.888799
S. Tsuruta, K. Ohsugi, Y. Toyama, T. Onoyama
To overcome the difficulties in the successful expertise of postal delivery and to make postal delivery more efficient and safe, a hybrid method for postal delivery route generation, integrating expertise with general algorithms, was proposed. Owing to this method, our system is able to automatically generate efficient delivery routes, which are also safe, considerate of the residents and their environment, and/or respect precedents or previous routes. Thus, this system can be used in various areas, from over-populated super-cities to conventional countryside areas. Furthermore, in the proposed method, the knowledge processing part and the general algorithm part are connected only by data called a "conceptual map", and the latter part is independent of domain knowledge that differs depending on experts, areas and time. Therefore, only the AI part needs modification in order to use this system in the various areas of offices or companies, and to use it for a long time.
{"title":"Hybrid method for a postal delivery route generation","authors":"S. Tsuruta, K. Ohsugi, Y. Toyama, T. Onoyama","doi":"10.1109/TENCON.2000.888799","DOIUrl":"https://doi.org/10.1109/TENCON.2000.888799","url":null,"abstract":"To overcome the difficulties in the successful expertise of postal delivery and to make postal delivery more efficient and safe, a hybrid method for postal delivery route generation, integrating expertise with general algorithms, was proposed. Owing to this method, our system is able to automatically generate efficient delivery routes, which are also safe, considerate of the residents and their environment, and/or respect precedents or previous routes. Thus, this system can be used in various areas, from over-populated super-cities to conventional countryside areas. Furthermore, in the proposed method, the knowledge processing part and the general algorithm part are connected only by data called a \"conceptual map\", and the latter part is independent of domain knowledge that differs depending on experts, areas and time. Therefore, only the AI part needs modification in order to use this system in the various areas of offices or companies, and to use it for a long time.","PeriodicalId":396649,"journal":{"name":"2000 TENCON Proceedings. Intelligent Systems and Technologies for the New Millennium (Cat. No.00CH37119)","volume":"550 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2000-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120975329","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}
Pub Date : 2000-09-24DOI: 10.1109/TENCON.2000.888802
S. Abidi, J. Ong
Self-organising neural networks have a natural propensity to cluster well-defined data into visually distinct clusters, which can then be easily interpretable by data analysts. However, there are situations when the clustering output of the self-organising network does not render distinct clusters. In this paper, we present a technique to automate the data mining task of data clustering, i.e. to automate cluster identification/demarcation by drawing upon a synergy between the self-organising neural networks and statistical data clustering techniques. The implied hybrid of diverse data clustering techniques provides an improved strategy to (a) discover hidden similarities between data items; (b) group similar data items into distinct and well-defined clusters - i.e. with explicit boundaries between different clusters and defined cluster membership characteristics; and (c) visualise the emergent data clusters in a 2D and 3D manner. Our proposed solution is implemented in terms of a data clustering workbench (DCW) - an all-encompassing (exploratory) data mining application.
{"title":"A data mining strategy for inductive data clustering: a synergy between self-organising neural networks and K-means clustering techniques","authors":"S. Abidi, J. Ong","doi":"10.1109/TENCON.2000.888802","DOIUrl":"https://doi.org/10.1109/TENCON.2000.888802","url":null,"abstract":"Self-organising neural networks have a natural propensity to cluster well-defined data into visually distinct clusters, which can then be easily interpretable by data analysts. However, there are situations when the clustering output of the self-organising network does not render distinct clusters. In this paper, we present a technique to automate the data mining task of data clustering, i.e. to automate cluster identification/demarcation by drawing upon a synergy between the self-organising neural networks and statistical data clustering techniques. The implied hybrid of diverse data clustering techniques provides an improved strategy to (a) discover hidden similarities between data items; (b) group similar data items into distinct and well-defined clusters - i.e. with explicit boundaries between different clusters and defined cluster membership characteristics; and (c) visualise the emergent data clusters in a 2D and 3D manner. Our proposed solution is implemented in terms of a data clustering workbench (DCW) - an all-encompassing (exploratory) data mining application.","PeriodicalId":396649,"journal":{"name":"2000 TENCON Proceedings. Intelligent Systems and Technologies for the New Millennium (Cat. No.00CH37119)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2000-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114912721","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}
Pub Date : 2000-09-24DOI: 10.1109/TENCON.2000.888798
H. Salleh, T.F. Yusaf, M. Azlan
This project is a result of a number of tests and experimental work that has been carried out at the UNITEN Laboratory to control the water level via the Internet (using LabVIEW/sup TM/). The experimental setup consists of a water tank, a sump, a level sensor, a proportional valve and a centrifugal pump. The computer data acquisition and control system uses a plug-in module which consists of a converter, a digital input/output module, an analog output module and an analog input module. This project also represents the first phase of controlling and operating a stationary single-cylinder four-stroke diesel engine.
{"title":"Level control experiment via Internet","authors":"H. Salleh, T.F. Yusaf, M. Azlan","doi":"10.1109/TENCON.2000.888798","DOIUrl":"https://doi.org/10.1109/TENCON.2000.888798","url":null,"abstract":"This project is a result of a number of tests and experimental work that has been carried out at the UNITEN Laboratory to control the water level via the Internet (using LabVIEW/sup TM/). The experimental setup consists of a water tank, a sump, a level sensor, a proportional valve and a centrifugal pump. The computer data acquisition and control system uses a plug-in module which consists of a converter, a digital input/output module, an analog output module and an analog input module. This project also represents the first phase of controlling and operating a stationary single-cylinder four-stroke diesel engine.","PeriodicalId":396649,"journal":{"name":"2000 TENCON Proceedings. Intelligent Systems and Technologies for the New Millennium (Cat. No.00CH37119)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2000-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114926371","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}
Pub Date : 2000-09-24DOI: 10.1109/TENCON.2000.888816
H. Saim, S. Fhong, N. Noor, Junaidy Bin Abd Wahab
The speckle pattern produced by the use of a coherent transducer in forming the B-mode ultrasound images is also a source of contrast resolution-degradation. We have presented the soft thresholding rule to relieve artifacts from the discrete wavelet transform of an image. Our study begins with the estimation of the required determinants to obtain the desired noise/resolution trade off, extended to the gray level mapping technique (GLM) application. The GLM application is used to further enhance the ultrasound images in different contrast levels. Experiments have shown that the proposed methods have significantly improved PSNR in the specular and reflectance area. Both techniques are translated into a PC-based tool that could be transferred into ultrasound system that can be implemented for real time image processing.
{"title":"Contrast resolution enhancement based on wavelet shrinkage and gray level mapping technique","authors":"H. Saim, S. Fhong, N. Noor, Junaidy Bin Abd Wahab","doi":"10.1109/TENCON.2000.888816","DOIUrl":"https://doi.org/10.1109/TENCON.2000.888816","url":null,"abstract":"The speckle pattern produced by the use of a coherent transducer in forming the B-mode ultrasound images is also a source of contrast resolution-degradation. We have presented the soft thresholding rule to relieve artifacts from the discrete wavelet transform of an image. Our study begins with the estimation of the required determinants to obtain the desired noise/resolution trade off, extended to the gray level mapping technique (GLM) application. The GLM application is used to further enhance the ultrasound images in different contrast levels. Experiments have shown that the proposed methods have significantly improved PSNR in the specular and reflectance area. Both techniques are translated into a PC-based tool that could be transferred into ultrasound system that can be implemented for real time image processing.","PeriodicalId":396649,"journal":{"name":"2000 TENCON Proceedings. Intelligent Systems and Technologies for the New Millennium (Cat. No.00CH37119)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2000-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115436199","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：