Pub Date : 2001-07-15DOI: 10.1109/IJCNN.2001.939487
G. Castellano, A. Fanelli, T. Roselli
Addresses the problem of user modeling, which is a crucial step in the development of adaptive hypermedia systems. In particular, we focus on adaptive educational hypermedia systems, where the users are learners. Learners are modeled in the form of categories that are extracted from empirical data, represented by responses to questionnaires, via a competitive neural network. The key feature of the proposed network is that it is able to adapt its structure during learning so that the appropriate number of categories is automatically revealed. The effectiveness of the proposed approach is shown on two questionnaires of different type.
{"title":"Mining categories of learners by a competitive neural network","authors":"G. Castellano, A. Fanelli, T. Roselli","doi":"10.1109/IJCNN.2001.939487","DOIUrl":"https://doi.org/10.1109/IJCNN.2001.939487","url":null,"abstract":"Addresses the problem of user modeling, which is a crucial step in the development of adaptive hypermedia systems. In particular, we focus on adaptive educational hypermedia systems, where the users are learners. Learners are modeled in the form of categories that are extracted from empirical data, represented by responses to questionnaires, via a competitive neural network. The key feature of the proposed network is that it is able to adapt its structure during learning so that the appropriate number of categories is automatically revealed. The effectiveness of the proposed approach is shown on two questionnaires of different type.","PeriodicalId":346955,"journal":{"name":"IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2001-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116437927","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 : 2001-07-15DOI: 10.1109/IJCNN.2001.939519
I. Valova, N. Georgieva, Y. Kosugi
Our aim is to simulate the dynamic behavior of the olfactory bulb as part of the olfactory system. Olfactory EEG have revealed that oscillation and chaos play important roles in the processing of information in the bulb. We have based our model on coupled nonlinear oscillators, which resemble groups of mitral and granule cells as main building units. The model involves excitatory mitral and inhibitory granule cells, forming a non-linear oscillator. Several of these oscillators are coupled in a two layer architecture. The system exhibits complex oscillatory behavior, simulating the mammalian olfactory bulb. Results for two different types of input are considered. Simulations show that the dynamic behavior of the model is stable under the influence of noise. The model bulb responds to different odor input with spatio-temporal activation patterns, which are unique for each simulated odor. After inhalation has started, a burst of oscillatory activity emerges. The specific pattern of oscillation, which is exhibited by the bulb model, is coherent over the whole bulb.
{"title":"Modeling of the odor information processing in the mammalian brain","authors":"I. Valova, N. Georgieva, Y. Kosugi","doi":"10.1109/IJCNN.2001.939519","DOIUrl":"https://doi.org/10.1109/IJCNN.2001.939519","url":null,"abstract":"Our aim is to simulate the dynamic behavior of the olfactory bulb as part of the olfactory system. Olfactory EEG have revealed that oscillation and chaos play important roles in the processing of information in the bulb. We have based our model on coupled nonlinear oscillators, which resemble groups of mitral and granule cells as main building units. The model involves excitatory mitral and inhibitory granule cells, forming a non-linear oscillator. Several of these oscillators are coupled in a two layer architecture. The system exhibits complex oscillatory behavior, simulating the mammalian olfactory bulb. Results for two different types of input are considered. Simulations show that the dynamic behavior of the model is stable under the influence of noise. The model bulb responds to different odor input with spatio-temporal activation patterns, which are unique for each simulated odor. After inhalation has started, a burst of oscillatory activity emerges. The specific pattern of oscillation, which is exhibited by the bulb model, is coherent over the whole bulb.","PeriodicalId":346955,"journal":{"name":"IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222)","volume":"87 3-4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2001-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123432927","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 : 2001-07-15DOI: 10.1109/IJCNN.2001.938792
M. Siddique, M. Tokhi
There are a number of problems associated with training neural networks with backpropagation algorithm. The algorithm scales exponentially with increased complexity of the problem. It is very often trapped in local minima, and is not robust to changes of network parameters such as number of hidden layer neurons and learning rate. The use of genetic algorithms is a recent trend, which is good at exploring a large and complex search space, to overcome such problems. In this paper a genetic algorithm is proposed for training feedforward neural networks and its performances is investigated. The results are analyzed and compared with those obtained by the backpropagation algorithm.
{"title":"Training neural networks: backpropagation vs. genetic algorithms","authors":"M. Siddique, M. Tokhi","doi":"10.1109/IJCNN.2001.938792","DOIUrl":"https://doi.org/10.1109/IJCNN.2001.938792","url":null,"abstract":"There are a number of problems associated with training neural networks with backpropagation algorithm. The algorithm scales exponentially with increased complexity of the problem. It is very often trapped in local minima, and is not robust to changes of network parameters such as number of hidden layer neurons and learning rate. The use of genetic algorithms is a recent trend, which is good at exploring a large and complex search space, to overcome such problems. In this paper a genetic algorithm is proposed for training feedforward neural networks and its performances is investigated. The results are analyzed and compared with those obtained by the backpropagation algorithm.","PeriodicalId":346955,"journal":{"name":"IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222)","volume":"66 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2001-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124798997","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 : 2001-07-15DOI: 10.1109/IJCNN.2001.939086
H. El-Bakry
A combination of fast and cooperative modular neural nets to enhance the performance of the detection process is introduced. I have applied such an approach successfully to detect human faces in cluttered scenes (El-Bakry et al.) (2000). Here, this technique is used to identify human irises automatically in a given image. In the detection phase, neural nets are used to test whether a window of 20/spl times/20 pixels contains an iris or not. The major difficulty in the learning process comes from the large database required for iris/non-iris images. A simple design for cooperative modular neural nets is presented to solve this problem by dividing these data into three groups. Such division results in reduction of computational complexity and thus decreasing the time and memory needed during the test of an image. Simulation results for the proposed algorithm show a good performance. Furthermore, faster iris detection is obtained through image decomposition into many sub-images and applying cross correlation in the frequency domain between each sub-image and the weights of the hidden layer.
介绍了一种结合快速和协作的模块化神经网络来提高检测过程的性能。我已经成功地应用了这种方法来检测混乱场景中的人脸(El-Bakry et al.)(2000)。在这里，这项技术被用来在给定的图像中自动识别人类的虹膜。在检测阶段，使用神经网络测试20/spl次/20像素的窗口是否包含虹膜。学习过程中的主要困难来自虹膜/非虹膜图像所需的大型数据库。提出了一种简单的协作模块化神经网络设计，通过将这些数据分成三组来解决这一问题。这样的分割结果降低了计算复杂度，从而减少了图像测试期间所需的时间和内存。仿真结果表明，该算法具有良好的性能。此外，通过将图像分解成许多子图像，并在每个子图像与隐藏层权值之间进行频域相互关联，从而获得更快的虹膜检测速度。
{"title":"Human iris detection using fast cooperative modular neural nets","authors":"H. El-Bakry","doi":"10.1109/IJCNN.2001.939086","DOIUrl":"https://doi.org/10.1109/IJCNN.2001.939086","url":null,"abstract":"A combination of fast and cooperative modular neural nets to enhance the performance of the detection process is introduced. I have applied such an approach successfully to detect human faces in cluttered scenes (El-Bakry et al.) (2000). Here, this technique is used to identify human irises automatically in a given image. In the detection phase, neural nets are used to test whether a window of 20/spl times/20 pixels contains an iris or not. The major difficulty in the learning process comes from the large database required for iris/non-iris images. A simple design for cooperative modular neural nets is presented to solve this problem by dividing these data into three groups. Such division results in reduction of computational complexity and thus decreasing the time and memory needed during the test of an image. Simulation results for the proposed algorithm show a good performance. Furthermore, faster iris detection is obtained through image decomposition into many sub-images and applying cross correlation in the frequency domain between each sub-image and the weights of the hidden layer.","PeriodicalId":346955,"journal":{"name":"IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2001-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129726990","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 : 2001-07-15DOI: 10.1109/IJCNN.2001.939565
V. Krasnopolsky
Two generic satellite remote sensing NN applications are described: NN solutions for forward and inverse (or retrieval) problems in satellite remote sensing. These two solutions correspond to two different approaches in satellite retrievals: variational retrievals (retrievals through the direct assimilation of sensor measurements) and standard retrievals. It is shown that both the forward model and the retrieval problem can be considered as nonlinear continuous mappings. The NN technique is a generic technique to perform continuous mappings. It is compared with regression approaches. Examples of a NN SSM/I forward model and a NN SSIM/I retrieval algorithm are used to illustrate advantages of using neural networks for developing both retrieval algorithms and forward models, and for minimizing the retrieval errors.
{"title":"Artificial neural networks in environmental sciences. I. NNs in satellite remote sensing and satellite meteorology","authors":"V. Krasnopolsky","doi":"10.1109/IJCNN.2001.939565","DOIUrl":"https://doi.org/10.1109/IJCNN.2001.939565","url":null,"abstract":"Two generic satellite remote sensing NN applications are described: NN solutions for forward and inverse (or retrieval) problems in satellite remote sensing. These two solutions correspond to two different approaches in satellite retrievals: variational retrievals (retrievals through the direct assimilation of sensor measurements) and standard retrievals. It is shown that both the forward model and the retrieval problem can be considered as nonlinear continuous mappings. The NN technique is a generic technique to perform continuous mappings. It is compared with regression approaches. Examples of a NN SSM/I forward model and a NN SSIM/I retrieval algorithm are used to illustrate advantages of using neural networks for developing both retrieval algorithms and forward models, and for minimizing the retrieval errors.","PeriodicalId":346955,"journal":{"name":"IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2001-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128453655","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 : 2001-07-15DOI: 10.1109/IJCNN.2001.938412
D. V. Prokhorov, L. Feldkamp, T. Feldkamp
We discuss an approach to joint density estimation called cluster-weighted modeling (CWM). The base approach was originally proposed by Gershenfeld (1998). We describe two innovations to the base CWM. Among these, the first enables the CWM to work with continuous streams of data. The second addresses the commonplace problem of local minima which may be encountered during the CWM parameter adjustment process. Our approach to mitigate this problem is quite elaborate, but it represents a principled way of improving the efficacy of the parameter adjustment process. We illustrate CWM and our performance enhancements with an example.
{"title":"A new approach to cluster-weighted modeling","authors":"D. V. Prokhorov, L. Feldkamp, T. Feldkamp","doi":"10.1109/IJCNN.2001.938412","DOIUrl":"https://doi.org/10.1109/IJCNN.2001.938412","url":null,"abstract":"We discuss an approach to joint density estimation called cluster-weighted modeling (CWM). The base approach was originally proposed by Gershenfeld (1998). We describe two innovations to the base CWM. Among these, the first enables the CWM to work with continuous streams of data. The second addresses the commonplace problem of local minima which may be encountered during the CWM parameter adjustment process. Our approach to mitigate this problem is quite elaborate, but it represents a principled way of improving the efficacy of the parameter adjustment process. We illustrate CWM and our performance enhancements with an example.","PeriodicalId":346955,"journal":{"name":"IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2001-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128592832","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 : 2001-07-15DOI: 10.1109/IJCNN.2001.939490
T. Akuzawa
Optimization problems on the general Lie group GL(N, |R) are naturally considered as those on the coset R/sup x(N)//GL(N, R) when the optimum is scale invariant. In this paper, we propose a new algorithm for optimization problems on this coset, named nested Newton's method, where we decompose the flow of optimization into quantum-like dynamics of N-particles under two-body interactions. Next, we propose a post-processing for independent component analysis (ICA) without pre-whitening, which we name the "post factor analysis" (post-FA). By post-FA we can estimate the noise variance beyond the known bound for the FA.
{"title":"Post factor analysis as a post-processing for ICA and new optimization algorithm as para-quantum dynamics","authors":"T. Akuzawa","doi":"10.1109/IJCNN.2001.939490","DOIUrl":"https://doi.org/10.1109/IJCNN.2001.939490","url":null,"abstract":"Optimization problems on the general Lie group GL(N, |R) are naturally considered as those on the coset R/sup x(N)//GL(N, R) when the optimum is scale invariant. In this paper, we propose a new algorithm for optimization problems on this coset, named nested Newton's method, where we decompose the flow of optimization into quantum-like dynamics of N-particles under two-body interactions. Next, we propose a post-processing for independent component analysis (ICA) without pre-whitening, which we name the \"post factor analysis\" (post-FA). By post-FA we can estimate the noise variance beyond the known bound for the FA.","PeriodicalId":346955,"journal":{"name":"IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222)","volume":"91 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2001-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128644857","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 : 2001-07-15DOI: 10.1109/IJCNN.2001.939508
A. Weitzenfeld, F. Cervantes, R. Sigala
Through experimentation and simulation scientists are able to get an understanding of the underlying biological mechanisms involved in living organisms. These mechanisms, both structural and behavioral, serve as inspiration in the modeling of neural based architectures as well as in the implementation of robotic systems. Among these, we are particularly motivated in studying animals such as toads, frogs, salamanders and praying mantis that rely on visuomotor coordination. In order to deal with the underlying complexity of these systems, we have developed the NSL/ASL simulation system to enable modeling and simulation at different levels of granularity.
{"title":"NSL/ASL: simulation of neural based visuomotor systems","authors":"A. Weitzenfeld, F. Cervantes, R. Sigala","doi":"10.1109/IJCNN.2001.939508","DOIUrl":"https://doi.org/10.1109/IJCNN.2001.939508","url":null,"abstract":"Through experimentation and simulation scientists are able to get an understanding of the underlying biological mechanisms involved in living organisms. These mechanisms, both structural and behavioral, serve as inspiration in the modeling of neural based architectures as well as in the implementation of robotic systems. Among these, we are particularly motivated in studying animals such as toads, frogs, salamanders and praying mantis that rely on visuomotor coordination. In order to deal with the underlying complexity of these systems, we have developed the NSL/ASL simulation system to enable modeling and simulation at different levels of granularity.","PeriodicalId":346955,"journal":{"name":"IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2001-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130131919","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 : 2001-07-15DOI: 10.1109/IJCNN.2001.939564
Isao Ha Yashi, J. Williamson
The topographic attentive mapping network based on a biologically-motivated neural network model is an especially effective model. When the network makes an incorrect output prediction, the attentional feedback circuit modulates the learning rates and adds a node to the category layer in order to improve the network's prediction accuracy. In this paper, a pruning method for reducing the number of category and feature nodes is formulated. We discuss the formulation and show its usefulness through some examples.
{"title":"Acquisition of fuzzy knowledge from topographic mixture networks with attentional feedback","authors":"Isao Ha Yashi, J. Williamson","doi":"10.1109/IJCNN.2001.939564","DOIUrl":"https://doi.org/10.1109/IJCNN.2001.939564","url":null,"abstract":"The topographic attentive mapping network based on a biologically-motivated neural network model is an especially effective model. When the network makes an incorrect output prediction, the attentional feedback circuit modulates the learning rates and adds a node to the category layer in order to improve the network's prediction accuracy. In this paper, a pruning method for reducing the number of category and feature nodes is formulated. We discuss the formulation and show its usefulness through some examples.","PeriodicalId":346955,"journal":{"name":"IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2001-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131055435","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 : 2001-07-15DOI: 10.1109/IJCNN.2001.939064
A. Lupini, F. Piazza, A. Uncini
An architecture for multichannel blind deconvolution in the frequency domain is presented. It is based on a complex-domain non-linear function, built with two spline functions, one for the real and one for the imaginary part, whose control points are adaptively changed using gradient-based techniques. B-spline functions are used since they allow us to impose only simple constraints on the control parameters in order to ensure the needed monotonously increasing characteristic. In the paper the adaptation rules for both the un-mixing matrix and the spline control points are also derived. Some experimental results that demonstrate the effectiveness of the proposed method are presented.
{"title":"Frequency domain multichannel blind deconvolution using adaptive spline functions","authors":"A. Lupini, F. Piazza, A. Uncini","doi":"10.1109/IJCNN.2001.939064","DOIUrl":"https://doi.org/10.1109/IJCNN.2001.939064","url":null,"abstract":"An architecture for multichannel blind deconvolution in the frequency domain is presented. It is based on a complex-domain non-linear function, built with two spline functions, one for the real and one for the imaginary part, whose control points are adaptively changed using gradient-based techniques. B-spline functions are used since they allow us to impose only simple constraints on the control parameters in order to ensure the needed monotonously increasing characteristic. In the paper the adaptation rules for both the un-mixing matrix and the spline control points are also derived. Some experimental results that demonstrate the effectiveness of the proposed method are presented.","PeriodicalId":346955,"journal":{"name":"IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2001-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122369531","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：