Joonwhoan Lee, Young-Min Cheon, Soon-Young Kim, Eun-Jong Park
If the emotion that a man or woman feels seeing color patterns in average sense can be extracted as rules, the result is useful to make an emotion-based color image retrieval system. This paper shows that the rough set theory provides a convenient tool for the purpose. We collect the emotion data when people see a set of predesigned random color patterns and extract the coarse rules for the emotional evaluation of the color patterns using VPRS (variable precision rough set) theory. Those rules can be used not only to approximately evaluate color patterns such as wall papers but also to set the initial conditions for the precise mapping system based on adaptive fuzzy logic from image features to emotion spaces represented by linguistic image scales.
{"title":"Emotional Evaluation of Color Patterns Based on Rough Sets","authors":"Joonwhoan Lee, Young-Min Cheon, Soon-Young Kim, Eun-Jong Park","doi":"10.1109/ISITC.2007.44","DOIUrl":"https://doi.org/10.1109/ISITC.2007.44","url":null,"abstract":"If the emotion that a man or woman feels seeing color patterns in average sense can be extracted as rules, the result is useful to make an emotion-based color image retrieval system. This paper shows that the rough set theory provides a convenient tool for the purpose. We collect the emotion data when people see a set of predesigned random color patterns and extract the coarse rules for the emotional evaluation of the color patterns using VPRS (variable precision rough set) theory. Those rules can be used not only to approximately evaluate color patterns such as wall papers but also to set the initial conditions for the precise mapping system based on adaptive fuzzy logic from image features to emotion spaces represented by linguistic image scales.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-11-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121433351","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}
Ridge functions are multivariate functions of the form g(a ldr x), where g is a univariate function, and a ldr x is the inner product of a isin Rd{0} and x isin Rd. We are concerned with the uniqueness of representation of a given function as some sum of ridge functions. We prove that if f(x) = Sigmai=1m gi(aildr x) = 0 for some ai = (a1i, hellip , adi) isin Rd{0}, and if gi isin Llocp(R) (or gi isin D' (R) and gi(ai ldr x) isin D' (Rd)), then, each gi is a polynomial of degree at most m - 2. We also prove a theorem on the smoothness of linear combinations of ridge functions. These results improve the existing results.
岭函数是形式为g(a ldr x)的多元函数，其中g是单变量函数，而a ldr x是a isin Rd{0}与x isin Rd的内积。我们关注的是给定函数表示为岭函数的一些和的唯一性。我们证明了如果f(x) = Sigmai=1 m gi(aildr x) = 0，对于某些ai= (a1 i, hellip, ad i) isin Rd{0}，且gi = Lloc p(R)(或gi = D' (R)和gi(aildr x) isin D' (Rd))，则每个gi都是至多m - 2次的多项式。我们还证明了脊函数线性组合的光滑性定理。这些结果改进了现有的结果。
{"title":"Uniqueness of Linear Combinations of Ridge Functions","authors":"Jinling Long, Wei Wu, Dong Nan, Junfang Wang","doi":"10.1109/ICNC.2007.790","DOIUrl":"https://doi.org/10.1109/ICNC.2007.790","url":null,"abstract":"Ridge functions are multivariate functions of the form g(a ldr x), where g is a univariate function, and a ldr x is the inner product of a isin R<sup>d</sup>{0} and x isin R<sup>d</sup>. We are concerned with the uniqueness of representation of a given function as some sum of ridge functions. We prove that if f(x) = Sigma<sub>i=1</sub> <sup>m</sup> g<sub>i</sub>(a<sup>i</sup>ldr x) = 0 for some a<sup>i</sup> = (a<sub>1</sub> <sup>i</sup>, hellip , a<sub>d</sub> <sup>i</sup>) isin R<sup>d</sup>{0}, and if g<sub>i</sub> isin L<sub>loc</sub> <sup>p</sup>(R) (or g<sub>i</sub> isin D' (R) and g<sub>i</sub>(a<sup>i</sup> ldr x) isin D' (R<sup>d</sup>)), then, each g<sub>i</sub> is a polynomial of degree at most m - 2. We also prove a theorem on the smoothness of linear combinations of ridge functions. These results improve the existing results.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129353477","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 : 2007-10-29DOI: 10.1109/ICMLC.2007.4370195
Huailin Shu, Huailin Shu
PIDNN (proportional, integral and derivative neural network) was first created by the author in 1997. In this paper, the author analyzes the characteristics of the temperature system of the plastic injecting-moulding machine and the performances of the PIDNN control system. The simulation results for the three-stage heater in a plastic injection machine are shown. It is proved that the PID neural network has perfect decoupling and self-learning control performances.
{"title":"PID Neural Network Temperature Control System in Plastic Injecting-moulding Machine","authors":"Huailin Shu, Huailin Shu","doi":"10.1109/ICMLC.2007.4370195","DOIUrl":"https://doi.org/10.1109/ICMLC.2007.4370195","url":null,"abstract":"PIDNN (proportional, integral and derivative neural network) was first created by the author in 1997. In this paper, the author analyzes the characteristics of the temperature system of the plastic injecting-moulding machine and the performances of the PIDNN control system. The simulation results for the three-stage heater in a plastic injection machine are shown. It is proved that the PID neural network has perfect decoupling and self-learning control performances.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123813858","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 membrane bioreactor (MBR) is a new technology of sewage treatment combining the membrane with the bioreactor, but the membrane fouling is an important factor to limit the MBR further development. Considering the issues that the relationship between the membrane fouling and affecting factors is a complicated and nonlinear, a modeling method based on wavelet neural network is presented. We adopt a method of reduce the number of the wavelet basic function by analysis the sparsity property of sample data, and use the learning algorithm based on gradient descent to train network. The main parameters of affecting MBR membrane fouling are studied. With the ability of strong function approach and fast convergence of wavelet network, the modeling method can detect and assess the membrane fouling degree of MBR in real time by learning the membrane fouling information. The detection results show that this method is feasible and effective.
{"title":"The Study of Membrane Fouling Modeling Method Based on Wavelet Neural Network for Sewage Treatment Membrane Bioreactor","authors":"Meyuan Gao, Jingwen Tian, Lixin Zhao, Kai Li","doi":"10.1109/ICICIC.2007.591","DOIUrl":"https://doi.org/10.1109/ICICIC.2007.591","url":null,"abstract":"The membrane bioreactor (MBR) is a new technology of sewage treatment combining the membrane with the bioreactor, but the membrane fouling is an important factor to limit the MBR further development. Considering the issues that the relationship between the membrane fouling and affecting factors is a complicated and nonlinear, a modeling method based on wavelet neural network is presented. We adopt a method of reduce the number of the wavelet basic function by analysis the sparsity property of sample data, and use the learning algorithm based on gradient descent to train network. The main parameters of affecting MBR membrane fouling are studied. With the ability of strong function approach and fast convergence of wavelet network, the modeling method can detect and assess the membrane fouling degree of MBR in real time by learning the membrane fouling information. The detection results show that this method is feasible and effective.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"2663 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126592551","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}
This paper proposes an improved particle swarm optimization algorithm(IPSO). IPSO adopts a new mutation operator and a new method that congregates some neighboring individuals to form multiple sub- populations in order to lead particles to explore new search space. Additionally, our algorithm incorporates a mechanism with a simple and easy penalty function to handle constraint. Thus, our algorithm has strong global exploratory capability and efficiency while being applied to solve nonlinear constrained optimization problems. Experimental results indicate that our IPSO is robust and efficient in solving nonlinear constrained optimization problems.
{"title":"An Improved Particle Swarm Algorithm for Solving Nonlinear Constrained Optimization Problems","authors":"Jinhua Zheng, Qian Wu, Wu Song","doi":"10.1109/ICNC.2007.221","DOIUrl":"https://doi.org/10.1109/ICNC.2007.221","url":null,"abstract":"This paper proposes an improved particle swarm optimization algorithm(IPSO). IPSO adopts a new mutation operator and a new method that congregates some neighboring individuals to form multiple sub- populations in order to lead particles to explore new search space. Additionally, our algorithm incorporates a mechanism with a simple and easy penalty function to handle constraint. Thus, our algorithm has strong global exploratory capability and efficiency while being applied to solve nonlinear constrained optimization problems. Experimental results indicate that our IPSO is robust and efficient in solving nonlinear constrained optimization problems.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115406084","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}
In this paper, facet analyses are made about the population sizing and sampling of the factorization based on Gaussian probability density function in the real- coded ECGA (rECGA) on the univariate and multivariate real-valued deceptive functions (URDF and MRDFi). The dynamics of the rECGA with single Gaussian pdf and mixture Gaussian pdf are described statistically. Experimental results illustrate that the rECGA with mixture Gaussian pdf has a scalability of sub-quadratic polynomial on the MRDFi, which indicates that it is applicable to large-scale decomposable optimization problems.
{"title":"Performance Analyses of Factorization Based on Gaussian PDF In rECGA","authors":"Minqiang Li, D. Goldberg, K. Sastry, Tian-Li Yu","doi":"10.1109/ICNC.2007.548","DOIUrl":"https://doi.org/10.1109/ICNC.2007.548","url":null,"abstract":"In this paper, facet analyses are made about the population sizing and sampling of the factorization based on Gaussian probability density function in the real- coded ECGA (rECGA) on the univariate and multivariate real-valued deceptive functions (URDF and MRDFi). The dynamics of the rECGA with single Gaussian pdf and mixture Gaussian pdf are described statistically. Experimental results illustrate that the rECGA with mixture Gaussian pdf has a scalability of sub-quadratic polynomial on the MRDFi, which indicates that it is applicable to large-scale decomposable optimization problems.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115699171","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}
Monetary policy affects the economy with long and variable lags, and for this reason policy-makers require reliable forecasts of economic activity. Hence, forecasts of real GDP growth have become more and more necessary. Haiming Guo (2006) proposed a new modified ARIMA model and used it to forecast the GDP growth of China from 1978 to 2004. Their experimental data show that the modified ARIMA model could provide more accurate forecasts than conventional ARIMA. However, all these models are linear. In this paper, we propose a new genetic programming method to forecast the GDP time series of China, United States and Japan from 1980 to 2006. Experimental results show that genetic programming yield statistically lower forecast errors for the year- over-year GDP data relative to modified linear ARIMA models. Moreover, we use the proposed method to forecast the future GDP growth of China, United States and Japan from 2007 to 2020, and we surprisingly find that the GDP of Japan fluctuates periodically, however the GDP of China and United States increases stably in the near future. According to the predicted data we can see that the GDP of China will exceed the GDP of Japan for the first time in 2020 or so.
{"title":"Forecasting GDP Growth Using Genetic Programming","authors":"Meifang Li, Guoxin Liu, Yongxiang Zhao","doi":"10.1109/ICNC.2007.388","DOIUrl":"https://doi.org/10.1109/ICNC.2007.388","url":null,"abstract":"Monetary policy affects the economy with long and variable lags, and for this reason policy-makers require reliable forecasts of economic activity. Hence, forecasts of real GDP growth have become more and more necessary. Haiming Guo (2006) proposed a new modified ARIMA model and used it to forecast the GDP growth of China from 1978 to 2004. Their experimental data show that the modified ARIMA model could provide more accurate forecasts than conventional ARIMA. However, all these models are linear. In this paper, we propose a new genetic programming method to forecast the GDP time series of China, United States and Japan from 1980 to 2006. Experimental results show that genetic programming yield statistically lower forecast errors for the year- over-year GDP data relative to modified linear ARIMA models. Moreover, we use the proposed method to forecast the future GDP growth of China, United States and Japan from 2007 to 2020, and we surprisingly find that the GDP of Japan fluctuates periodically, however the GDP of China and United States increases stably in the near future. According to the predicted data we can see that the GDP of China will exceed the GDP of Japan for the first time in 2020 or so.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"139 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124398548","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}
Support vector machine is intrinsically a binary classifier providing no theoretically formulated procedure for multi-category classification. Several methods have been developed to extend it to multi-category problems. Combining strengths of them, an improved "labeled multi-category support vector machine" is proposed. The proposed method explicitly labels samples and performs multi-category classification with only a single support vector machine classifier. Labeling samples leads to the sample number disparity between positive and negative classes. The techniques of setting different cost parameters for different classes are employed to enhance the algorithm's performance. Generalization error bound estimates are theoretically derived by the new technique of maximal discrepancy. Experiments with a benchmark dataset show that the algorithm can accurately classify multi-category data. Rotor mechanical fault recognition applications confirm that the algorithm can efficiently perform multi-category fault detection and identification.
{"title":"Fault Recognition with Labeled Multi-category Support Vector Machine","authors":"Xue Wang, Daowei Bi, Sheng Wang","doi":"10.1109/ICNC.2007.382","DOIUrl":"https://doi.org/10.1109/ICNC.2007.382","url":null,"abstract":"Support vector machine is intrinsically a binary classifier providing no theoretically formulated procedure for multi-category classification. Several methods have been developed to extend it to multi-category problems. Combining strengths of them, an improved \"labeled multi-category support vector machine\" is proposed. The proposed method explicitly labels samples and performs multi-category classification with only a single support vector machine classifier. Labeling samples leads to the sample number disparity between positive and negative classes. The techniques of setting different cost parameters for different classes are employed to enhance the algorithm's performance. Generalization error bound estimates are theoretically derived by the new technique of maximal discrepancy. Experiments with a benchmark dataset show that the algorithm can accurately classify multi-category data. Rotor mechanical fault recognition applications confirm that the algorithm can efficiently perform multi-category fault detection and identification.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116864633","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}
Aiming at denoising to detection signal of the flaw in the long transporting pipe, the way of denoising based on wavelet neural network is present, and signal processing of ultrasonic detection application in long pipeline is described. Making use of self-learning characteristic of wavelet neural network, this way reduces wave loss. This method has the good effect and may acquire exact location and amplitude of the flaw. It is great significance for signal processing of ultrasonic detection.
{"title":"Application of Signal Detection for Pipeline Flaw Based on Wavelet Neural Network","authors":"Runjing Zhou, Fei Zhang","doi":"10.1109/ICNC.2007.263","DOIUrl":"https://doi.org/10.1109/ICNC.2007.263","url":null,"abstract":"Aiming at denoising to detection signal of the flaw in the long transporting pipe, the way of denoising based on wavelet neural network is present, and signal processing of ultrasonic detection application in long pipeline is described. Making use of self-learning characteristic of wavelet neural network, this way reduces wave loss. This method has the good effect and may acquire exact location and amplitude of the flaw. It is great significance for signal processing of ultrasonic detection.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116993937","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 intelligent optimization approach is proposed for eigenstructure assignment (EA) via neural network (NN) adjusting the components of output vector autonomously. The basic idea is to minimize the L2 norm of error between the desired vector and achievable vector using the designing freedom provided by EA technique. Besides, close-loop eigenvalues are also optimised within desired regions on the left-half complex plane according to the design objective to ensure both closed-loop stability and dynamical performance. With the proposed approach, additional closed-loop specifications such as decoupling of different modes and robustness can also be easily achieved. As a demonstration, application of the proposed approach to the designing of flight control law for an advanced fighter is discussed. The simulation results show good closed loop performance and validate the proposed intelligent optimization approach of EA technique.
{"title":"Eigenstructure Assignment Based Flight Control for Advanced Fighter: An Optimization Based Approach","authors":"Yong Fan, Jihong Zhu, Chunning Yang, Zeng-qi Sun","doi":"10.1109/ICNC.2007.350","DOIUrl":"https://doi.org/10.1109/ICNC.2007.350","url":null,"abstract":"An intelligent optimization approach is proposed for eigenstructure assignment (EA) via neural network (NN) adjusting the components of output vector autonomously. The basic idea is to minimize the L2 norm of error between the desired vector and achievable vector using the designing freedom provided by EA technique. Besides, close-loop eigenvalues are also optimised within desired regions on the left-half complex plane according to the design objective to ensure both closed-loop stability and dynamical performance. With the proposed approach, additional closed-loop specifications such as decoupling of different modes and robustness can also be easily achieved. As a demonstration, application of the proposed approach to the designing of flight control law for an advanced fighter is discussed. The simulation results show good closed loop performance and validate the proposed intelligent optimization approach of EA technique.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"101 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2007-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117278644","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：