Precise description of the engine dynamic characteristics plays a crucial role in automatic gear-shifting decision making for the performance match and optimization of vehicle power-train system. In this paper, a multi-layer feed forward neural network was proposed to identify the dynamic torque and fuel consumption models of engine. Based on the neural network models, algorithms to calculate the optimal dynamic and economical gear-shifting rules were constructed respectively. Comparative tests show that the gear-shifting decision based on neural network computation models is better than that based on traditional computation model using curve approximation, and improves the dynamic performance and fuel economy of vehicle power-train system significantly.
{"title":"Automotive Gear-Shifting Decision Making Based on Neural Network Computation Model","authors":"Jingxing Tan, Xiaofeng Yin, Liang Yin, Ling Zhao","doi":"10.1109/ICNC.2007.279","DOIUrl":"https://doi.org/10.1109/ICNC.2007.279","url":null,"abstract":"Precise description of the engine dynamic characteristics plays a crucial role in automatic gear-shifting decision making for the performance match and optimization of vehicle power-train system. In this paper, a multi-layer feed forward neural network was proposed to identify the dynamic torque and fuel consumption models of engine. Based on the neural network models, algorithms to calculate the optimal dynamic and economical gear-shifting rules were constructed respectively. Comparative tests show that the gear-shifting decision based on neural network computation models is better than that based on traditional computation model using curve approximation, and improves the dynamic performance and fuel economy of vehicle power-train system significantly.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"227 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":"131934614","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}
Hongbin Suo, Ming Li, Tantan Liu, Ping Lu, Yonghong Yan
The design approach for classifying the backend features of the PPRLM (Parallel Phone Recognition and Language Modeling) system is demonstrated in this paper. A variety of features and their combinations extracted by language dependent recognizers were evaluated based on the National Institute of Standards and Technology (NIST) Language Recognition Evaluation (LRE) 2003 corpus. Three well-known classifiers: Gaussian Mixture Model (GMM), Support Vector Machine (SVM), and feed forward neural network (NN) are proposed to compartmentalize these high level features which are generated by n-gram language model scoring and one pass decoding based on acoustic model in PPRLM system. Finally, the log-likelihood radio (LLR) normalization is applied to backend processing to the target language scores and the performance of language recognition is enhanced.
{"title":"The Design of Backend Classifiers in PPRLM System for Language Identification","authors":"Hongbin Suo, Ming Li, Tantan Liu, Ping Lu, Yonghong Yan","doi":"10.1109/ICNC.2007.719","DOIUrl":"https://doi.org/10.1109/ICNC.2007.719","url":null,"abstract":"The design approach for classifying the backend features of the PPRLM (Parallel Phone Recognition and Language Modeling) system is demonstrated in this paper. A variety of features and their combinations extracted by language dependent recognizers were evaluated based on the National Institute of Standards and Technology (NIST) Language Recognition Evaluation (LRE) 2003 corpus. Three well-known classifiers: Gaussian Mixture Model (GMM), Support Vector Machine (SVM), and feed forward neural network (NN) are proposed to compartmentalize these high level features which are generated by n-gram language model scoring and one pass decoding based on acoustic model in PPRLM system. Finally, the log-likelihood radio (LLR) normalization is applied to backend processing to the target language scores and the performance of language recognition is enhanced.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"81 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":"132086257","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}
Motivated by the goal of discovering more accurate characteristics of Chinese stock market, this paper investigates the power-law properties and criticality of the Shanghai Securities Exchange Compound Index (SSEECI) with two benchmarks of 5-min and 1-day database. We find that the center profile of returns distribution is well described by Levy regime and, more important, that the approximately symmetric tails of distribution are characterized by another power-law regime with an exponent well out of Levy range 04days, the distribution exhibits the slow convergence to normal Gaussian behavior. The phenomena support that the critical timescale Deltatap4days of fully developed markets is universal for Chinese stock market.
{"title":"Multiscale Power-Law Properties and Criticality of Chinese Stock Market","authors":"Hong-lin Yang, Shou Chen, Yan Yang","doi":"10.1109/ICNC.2007.492","DOIUrl":"https://doi.org/10.1109/ICNC.2007.492","url":null,"abstract":"Motivated by the goal of discovering more accurate characteristics of Chinese stock market, this paper investigates the power-law properties and criticality of the Shanghai Securities Exchange Compound Index (SSEECI) with two benchmarks of 5-min and 1-day database. We find that the center profile of returns distribution is well described by Levy regime and, more important, that the approximately symmetric tails of distribution are characterized by another power-law regime with an exponent well out of Levy range 0<alpha<2 and also beyond the exponent alphaap3 of fully developed markets. Moreover, we also show that returns appear to exhibit the criticality. When timescale Deltat>4days, the distribution exhibits the slow convergence to normal Gaussian behavior. The phenomena support that the critical timescale Deltatap4days of fully developed markets is universal for Chinese stock market.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"29 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":"132169014","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 algorithm for seeking the shortest path between two nodes in city's road net according to the characteristics of the net. The algorithm takes advantage of the theories of bidirectional search, projection, minimum angle and binary tree. According to the algorithm analysis , the algorithm can reduce seeking space and raise seeking speed greatly, and its time complexity can not exceed 0(N), while N is the number of road network nodes. The application results show that the algorithm has good practicability.
{"title":"A Fast Path Planning Algorithm for Vehicle Navigation System","authors":"Bi Jun, Guang-yu Zhu, Zheng-yu Xie","doi":"10.1109/ICNC.2007.26","DOIUrl":"https://doi.org/10.1109/ICNC.2007.26","url":null,"abstract":"This paper proposes an algorithm for seeking the shortest path between two nodes in city's road net according to the characteristics of the net. The algorithm takes advantage of the theories of bidirectional search, projection, minimum angle and binary tree. According to the algorithm analysis , the algorithm can reduce seeking space and raise seeking speed greatly, and its time complexity can not exceed 0(N), while N is the number of road network nodes. The application results show that the algorithm has good practicability.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"11 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":"133233927","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}
sorting based algorithm is proposed in this paper for finding non-dominated set in Multi-Objective optimization. The algorithm is composed by sorting step and dominated solutions deleting step. Some enhancement techniques including primary non-dominated solutions, scoring and summation sequence are used to reduce the computa tional complexity. Compared with the classic Kung et al.'s efficient algorithm, experiments show sorting based algorithm performs almost the same efficiently as the Kung et al.'s algorithm when there are less objectives and solutions, and much better when there are more objectives and solutions.
提出了一种基于排序的多目标优化非支配集查找算法。该算法由排序步骤和劣势解删除步骤组成。采用主非支配解、计分和数列等增强技术来降低计算复杂度。与经典的Kung et al.高效算法相比，实验表明，当目标和解较少时，基于排序的算法与Kung et al.算法的效率基本相同，而当目标和解较多时，基于排序的算法的效率更高。
{"title":"A Sorting Based Algorithm for Finding a Non-dominated Set in Multi-objective Optimization","authors":"Jun Du, Z. Cai, Yunliang Chen","doi":"10.1109/ICNC.2007.142","DOIUrl":"https://doi.org/10.1109/ICNC.2007.142","url":null,"abstract":"sorting based algorithm is proposed in this paper for finding non-dominated set in Multi-Objective optimization. The algorithm is composed by sorting step and dominated solutions deleting step. Some enhancement techniques including primary non-dominated solutions, scoring and summation sequence are used to reduce the computa tional complexity. Compared with the classic Kung et al.'s efficient algorithm, experiments show sorting based algorithm performs almost the same efficiently as the Kung et al.'s algorithm when there are less objectives and solutions, and much better when there are more objectives and solutions.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"48 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":"127069458","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, a time series prediction method using clustering to improve neural network is studied. The big data group is divided into some small parts by clustering. By this way, every small part has a higher conformity, and data in these small parts is used to train corresponding neural network for prediction. The prediction model is constructed from neural network with the addition of clustering and is applied to the financial time series prediction. The experiment results demonstrate the effectiveness of the improvement. Comparison with the primitive neural network prediction model shows that clustering increases neural network's trend accuracy in continuous prediction, while debasing the cost of time and reducing the complexity of the prediction model.
{"title":"Use clustering to improve neural network in financial time series prediction","authors":"Fen Liu, Peng Du, Fangfei Weng, Jun Qu","doi":"10.1109/ICNC.2007.796","DOIUrl":"https://doi.org/10.1109/ICNC.2007.796","url":null,"abstract":"In this paper, a time series prediction method using clustering to improve neural network is studied. The big data group is divided into some small parts by clustering. By this way, every small part has a higher conformity, and data in these small parts is used to train corresponding neural network for prediction. The prediction model is constructed from neural network with the addition of clustering and is applied to the financial time series prediction. The experiment results demonstrate the effectiveness of the improvement. Comparison with the primitive neural network prediction model shows that clustering increases neural network's trend accuracy in continuous prediction, while debasing the cost of time and reducing the complexity of the prediction model.","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":"124350484","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}
Jingbo Zhang, X. Zeng, Yinghua Lu, Lei Zhang, Meng Li
This paper proposes a novel off-line signature verification method based on one-class-one-network classification, using four groups features. The features include direction features, texture features, dynamic features and complexity index. At last, one-class-one-network classifier is used to verify the signatures. The signature verification system was experimented on real data sets and the results show the system is effective with the average error rate can reach 1.8%, which is obviously satisfactory.
{"title":"A Novel Off-line Signature Verification Based on One-class-one-network","authors":"Jingbo Zhang, X. Zeng, Yinghua Lu, Lei Zhang, Meng Li","doi":"10.1109/ICNC.2007.118","DOIUrl":"https://doi.org/10.1109/ICNC.2007.118","url":null,"abstract":"This paper proposes a novel off-line signature verification method based on one-class-one-network classification, using four groups features. The features include direction features, texture features, dynamic features and complexity index. At last, one-class-one-network classifier is used to verify the signatures. The signature verification system was experimented on real data sets and the results show the system is effective with the average error rate can reach 1.8%, which is obviously satisfactory.","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":"124568649","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}
Particle swarm optimization (PSO) has been shown to be an efficient, robust and simple optimization algorithm. Aim at the shortcoming that the PSO algorithm falls into local optimization easily, in this paper fuzzy control theory is introduced into PSO (FPSO). Parameters may be dynamic adjusted themselves according to the optimization effect every time in this algorithm. Its ability of dynamic adjustment is strengthened, and the global optimization performance of the algorithm can be improved better. And in this paper, the improved algorithm is illustrated how could solve the problem, which exists in the raw material requirement model for production processing in the ore dressing plant. The experimental results are provided to support the conclusions drawn from the theoretical findings.
{"title":"Research on Particle Swarm Optimization and its Industrial Application","authors":"Xiaoling Huang, N. Sun, W. Liu, Junxiu Wei","doi":"10.1109/ICNC.2007.628","DOIUrl":"https://doi.org/10.1109/ICNC.2007.628","url":null,"abstract":"Particle swarm optimization (PSO) has been shown to be an efficient, robust and simple optimization algorithm. Aim at the shortcoming that the PSO algorithm falls into local optimization easily, in this paper fuzzy control theory is introduced into PSO (FPSO). Parameters may be dynamic adjusted themselves according to the optimization effect every time in this algorithm. Its ability of dynamic adjustment is strengthened, and the global optimization performance of the algorithm can be improved better. And in this paper, the improved algorithm is illustrated how could solve the problem, which exists in the raw material requirement model for production processing in the ore dressing plant. The experimental results are provided to support the conclusions drawn from the theoretical findings.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"5 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":"114364208","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}
Taking electrical signals in the chrysanthemum (Dendranthema morifolium) as the time series and using the Gaussian radial base function (RBF) and a delayed input window chosen at 50, an intelligent RBF forecast system is set up to forecast signals by the wavelet soft-threshold de-noised backward. It is obvious that the electrical signal in chrysanthemum is a sort of weak, unstable and low frequency signals. There is the maximum amplitude at 1093.44 muV, minimum -605.35 muV, average value -11.94 muV; and below 0.3 Hz at frequency in the chrysanthemum respectively. A result shows that it is feasible to forecast plant electrical signals for the timing by using of the RBF neural network. The forecast data can be used as the important preferences for the intelligent automatic control system based on the adaptive characteristic of plants to achieve the energy saving on the agricultural production in the greenhouse and/or the plastic lookum.
{"title":"Prediction to the Weak Electrical Signal in Chrysanthemum by RBF Neural Networks","authors":"Jinli Ding, Miao Wang, Lanzhou Wang, Qiao Li","doi":"10.1109/ICNC.2007.565","DOIUrl":"https://doi.org/10.1109/ICNC.2007.565","url":null,"abstract":"Taking electrical signals in the chrysanthemum (Dendranthema morifolium) as the time series and using the Gaussian radial base function (RBF) and a delayed input window chosen at 50, an intelligent RBF forecast system is set up to forecast signals by the wavelet soft-threshold de-noised backward. It is obvious that the electrical signal in chrysanthemum is a sort of weak, unstable and low frequency signals. There is the maximum amplitude at 1093.44 muV, minimum -605.35 muV, average value -11.94 muV; and below 0.3 Hz at frequency in the chrysanthemum respectively. A result shows that it is feasible to forecast plant electrical signals for the timing by using of the RBF neural network. The forecast data can be used as the important preferences for the intelligent automatic control system based on the adaptive characteristic of plants to achieve the energy saving on the agricultural production in the greenhouse and/or the plastic lookum.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"2 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":"114408869","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 existing approaches to automatic term recognition include these types: dictionary-based, rule-based, statistical, etc. First, we discuss the dictionary-based methods briefly in this paper. Then we propose an approach for Chinese single word term extraction combining the dictionary-based method with seed knowledge-based method. Our method is based on two resources. One is the Chinese concept dictionary which is a general bilingual semantic lexicon and the other one is the bilingual seeds set extracted from a bilingual glossary of HK law. The approach is to recognize the legal domain-specific term. Our approach is applying general semantic lexicon for domain-specific term extraction. The experimental results show that our approach can get high precision in legal field. Keywords: automatic term recognition, bilingual seeds set, Chinese concept dictionary, legal terminology, single word term.
{"title":"Single Word Term Extraction Using a Bilingual Semantic Lexicon-Based Approach","authors":"Hongying Zan, Guocheng Duan, Minghong Fan","doi":"10.1109/ICNC.2007.667","DOIUrl":"https://doi.org/10.1109/ICNC.2007.667","url":null,"abstract":"The existing approaches to automatic term recognition include these types: dictionary-based, rule-based, statistical, etc. First, we discuss the dictionary-based methods briefly in this paper. Then we propose an approach for Chinese single word term extraction combining the dictionary-based method with seed knowledge-based method. Our method is based on two resources. One is the Chinese concept dictionary which is a general bilingual semantic lexicon and the other one is the bilingual seeds set extracted from a bilingual glossary of HK law. The approach is to recognize the legal domain-specific term. Our approach is applying general semantic lexicon for domain-specific term extraction. The experimental results show that our approach can get high precision in legal field. Keywords: automatic term recognition, bilingual seeds set, Chinese concept dictionary, legal terminology, single word term.","PeriodicalId":250881,"journal":{"name":"Third International Conference on Natural Computation (ICNC 2007)","volume":"4 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":"114380076","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：