Pub Date : 2012-07-15DOI: 10.1109/ICMLC.2012.6358928
Changzhong Wang, Xin-Hua Cui, Wenying Bao, Qiang He
In this paper, we study attribute reduction of decision system based on similar relations. We first define the concept of attribute reduction. We then develop a sufficient and necessary condition for attribute reduction and construct the dicernibility matrix on similar relations, by which we can compute all the reducts of decision systems. The experimental results with UCI data sets show that the proposed reduction approach is an effective method to deal with numerical and categorical data sets.
{"title":"Attribute reduction of decision table based on similar relation","authors":"Changzhong Wang, Xin-Hua Cui, Wenying Bao, Qiang He","doi":"10.1109/ICMLC.2012.6358928","DOIUrl":"https://doi.org/10.1109/ICMLC.2012.6358928","url":null,"abstract":"In this paper, we study attribute reduction of decision system based on similar relations. We first define the concept of attribute reduction. We then develop a sufficient and necessary condition for attribute reduction and construct the dicernibility matrix on similar relations, by which we can compute all the reducts of decision systems. The experimental results with UCI data sets show that the proposed reduction approach is an effective method to deal with numerical and categorical data sets.","PeriodicalId":128006,"journal":{"name":"2012 International Conference on Machine Learning and Cybernetics","volume":"131 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115618290","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 : 2012-07-15DOI: 10.1109/ICMLC.2012.6359556
Jian-Guo Shi, Juan Wang, Meng Zhou, Bing Yin, Yan-Fang Shi
In this paper, we prove the endpoint estimates for the multilinear commutator related to the singular integral on the space of homogeneous type by means of the Lp (l <; p <; ∞) boundedness for the singular integral operator. The main tools of the proof are decomposition of function spaces and some general inequalities. The estimates we established are widely applied in support vector machines, especially the Minkowski's inequality and Holde's inequality which may have application to classification problems based on Kernel methods.
{"title":"Endpoint learning for multilinear commutator of singular integral on space of homogeneous type","authors":"Jian-Guo Shi, Juan Wang, Meng Zhou, Bing Yin, Yan-Fang Shi","doi":"10.1109/ICMLC.2012.6359556","DOIUrl":"https://doi.org/10.1109/ICMLC.2012.6359556","url":null,"abstract":"In this paper, we prove the endpoint estimates for the multilinear commutator related to the singular integral on the space of homogeneous type by means of the Lp (l <; p <; ∞) boundedness for the singular integral operator. The main tools of the proof are decomposition of function spaces and some general inequalities. The estimates we established are widely applied in support vector machines, especially the Minkowski's inequality and Holde's inequality which may have application to classification problems based on Kernel methods.","PeriodicalId":128006,"journal":{"name":"2012 International Conference on Machine Learning and Cybernetics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124193696","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 : 2012-07-15DOI: 10.1109/ICMLC.2012.6358884
Li-Juan Wang, Z. Hao
This paper proposes an improved link analysis based clustering ensemble method (ILCEM). ILCEM can transform binary data-cluster association matrix into real-valued matrix according to the similarity between clusters in all base clustering. The refined data-cluster association matrix can generate more information to clustering ensemble so as to improve the performance of clustering. Experimental results on three VCI datasets have shown that ILCEM is better than KMC, base clustering method and CSM+GKMC.
{"title":"An improved link analysis based clustering ensemble method","authors":"Li-Juan Wang, Z. Hao","doi":"10.1109/ICMLC.2012.6358884","DOIUrl":"https://doi.org/10.1109/ICMLC.2012.6358884","url":null,"abstract":"This paper proposes an improved link analysis based clustering ensemble method (ILCEM). ILCEM can transform binary data-cluster association matrix into real-valued matrix according to the similarity between clusters in all base clustering. The refined data-cluster association matrix can generate more information to clustering ensemble so as to improve the performance of clustering. Experimental results on three VCI datasets have shown that ILCEM is better than KMC, base clustering method and CSM+GKMC.","PeriodicalId":128006,"journal":{"name":"2012 International Conference on Machine Learning and Cybernetics","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114705078","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 : 2012-07-15DOI: 10.1109/ICMLC.2012.6359645
Yuh-Rau Wang, Wei-Hung Lin, Ling Yang
This paper proposed a novel digital signal processor (DSP ) based license plate detection algorithm. DSP is applied to achieve the real time detection. Kalman filter is used to promote the speed of license plate localization (LPL) and reduce the times of frames. Through our proposal the feature filter algorithm can enhance the feature of LP. Experimental results show that the method proposed can detect license plates and the accuracy is up to 99%.
{"title":"A novel DSP based real time license plate detection algorithm","authors":"Yuh-Rau Wang, Wei-Hung Lin, Ling Yang","doi":"10.1109/ICMLC.2012.6359645","DOIUrl":"https://doi.org/10.1109/ICMLC.2012.6359645","url":null,"abstract":"This paper proposed a novel digital signal processor (DSP ) based license plate detection algorithm. DSP is applied to achieve the real time detection. Kalman filter is used to promote the speed of license plate localization (LPL) and reduce the times of frames. Through our proposal the feature filter algorithm can enhance the feature of LP. Experimental results show that the method proposed can detect license plates and the accuracy is up to 99%.","PeriodicalId":128006,"journal":{"name":"2012 International Conference on Machine Learning and Cybernetics","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123665371","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 : 2012-07-15DOI: 10.1109/ICMLC.2012.6359650
G. Luh
Skin detection is the key technology in various image processing applications such as face detection. The aim of skin detection is to determine if a color pixel is a skin or non-skin color. Skin color is often considered to be a useful and discriminating image feature for facial area since it provides computationally effective yet, robust to variation in scale, orientation and partial occlusion. Nevertheless, skin detection is also an extremely challenging task since the skin color is sensitive to various factors such as illumination, ethnicity, individual characteristics and subject appearances. In this paper, an artificial immune network based skin detection scheme in several skin color spaces is proposed. Particle swarm optimization is employed to train/optimize skin/non-skin immune network classifiers. The performance of the method was evaluated employing images derived from the Internet.
{"title":"Skin color detection using artificial immune networks","authors":"G. Luh","doi":"10.1109/ICMLC.2012.6359650","DOIUrl":"https://doi.org/10.1109/ICMLC.2012.6359650","url":null,"abstract":"Skin detection is the key technology in various image processing applications such as face detection. The aim of skin detection is to determine if a color pixel is a skin or non-skin color. Skin color is often considered to be a useful and discriminating image feature for facial area since it provides computationally effective yet, robust to variation in scale, orientation and partial occlusion. Nevertheless, skin detection is also an extremely challenging task since the skin color is sensitive to various factors such as illumination, ethnicity, individual characteristics and subject appearances. In this paper, an artificial immune network based skin detection scheme in several skin color spaces is proposed. Particle swarm optimization is employed to train/optimize skin/non-skin immune network classifiers. The performance of the method was evaluated employing images derived from the Internet.","PeriodicalId":128006,"journal":{"name":"2012 International Conference on Machine Learning and Cybernetics","volume":"60 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126165092","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 : 2012-07-15DOI: 10.1109/ICMLC.2012.6359687
Jun Xu, Ruifeng Xu, Xiaolong Wang
This paper address the problem of identifying the sentiment polarity in financial news articles about a public company having potential effect on the future price of the company's stock. The problem is challenging due to the lack of reliable labeled training data and effective classification method. A feasible corpus building strategy is proposed and stock reviews are used for training, since the news polarity prediction is similar to the process of stock analyst drawing their conclusion by weighting the major event pros and cons of the company. The reviews can be annotated automatically by the grade given by the analyst. In addition, the consequent experiments also confirm it. Furthermore, we examine the effectiveness of using language modeling approaches to solve the sentiment classification of Chinese financial news articles. Two different approaches based on language model are employed and their comparisons with SVM and Naive Bayes are also performed in our research. The experiment results justify the effectiveness and robustness of the proposed language model approaches, which perform better than the approaches based on traditional machine learning techniques.
{"title":"Language model based Chinese financial news sentiment classification","authors":"Jun Xu, Ruifeng Xu, Xiaolong Wang","doi":"10.1109/ICMLC.2012.6359687","DOIUrl":"https://doi.org/10.1109/ICMLC.2012.6359687","url":null,"abstract":"This paper address the problem of identifying the sentiment polarity in financial news articles about a public company having potential effect on the future price of the company's stock. The problem is challenging due to the lack of reliable labeled training data and effective classification method. A feasible corpus building strategy is proposed and stock reviews are used for training, since the news polarity prediction is similar to the process of stock analyst drawing their conclusion by weighting the major event pros and cons of the company. The reviews can be annotated automatically by the grade given by the analyst. In addition, the consequent experiments also confirm it. Furthermore, we examine the effectiveness of using language modeling approaches to solve the sentiment classification of Chinese financial news articles. Two different approaches based on language model are employed and their comparisons with SVM and Naive Bayes are also performed in our research. The experiment results justify the effectiveness and robustness of the proposed language model approaches, which perform better than the approaches based on traditional machine learning techniques.","PeriodicalId":128006,"journal":{"name":"2012 International Conference on Machine Learning and Cybernetics","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129478304","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 : 2012-07-15DOI: 10.1109/ICMLC.2012.6359480
Yan Wang, Jing Fu, X. An, Jian Li, Er-Ke Shang
Vision-based Lane Departure Warning Systems (LDWSs) have been studied for over two decays. This paper presents an Objective Evaluation Platform of LDWS (OEP-LDWS). It provides simulated road scenes with the possible data as ground truth, such as the vehicle to road relation, vehicle's states and the real Time-to-Lane-Crossing (TLC) value. In our OEP-LDWS, different kinds of driving maneuver can be simulated with the road model, the vehicle model, the camera model and a vehicle trajectory generator. At the same time, the road scene that may be captured by the on board camera can be generated. Using our OEP-LDWS, one can not only evaluate the warning performance of the LDWS quantitatively, but also assess the whole performance under varying circumstance, such as different road surfaces, different road curvatures and so on. Actually, those assessments can hardly be evaluated through real driving test, and are very important aspects for a LDWS, such as warning strategy selection, system tailor. Using our OEP-LDWS, we assess our LDWS with three different warning strategies, and reach the conclusion that the PTLC is the best under the low false warning criterion.
{"title":"On the quantitative assessment of the Lane Departure Warning System based on road scenes simulator","authors":"Yan Wang, Jing Fu, X. An, Jian Li, Er-Ke Shang","doi":"10.1109/ICMLC.2012.6359480","DOIUrl":"https://doi.org/10.1109/ICMLC.2012.6359480","url":null,"abstract":"Vision-based Lane Departure Warning Systems (LDWSs) have been studied for over two decays. This paper presents an Objective Evaluation Platform of LDWS (OEP-LDWS). It provides simulated road scenes with the possible data as ground truth, such as the vehicle to road relation, vehicle's states and the real Time-to-Lane-Crossing (TLC) value. In our OEP-LDWS, different kinds of driving maneuver can be simulated with the road model, the vehicle model, the camera model and a vehicle trajectory generator. At the same time, the road scene that may be captured by the on board camera can be generated. Using our OEP-LDWS, one can not only evaluate the warning performance of the LDWS quantitatively, but also assess the whole performance under varying circumstance, such as different road surfaces, different road curvatures and so on. Actually, those assessments can hardly be evaluated through real driving test, and are very important aspects for a LDWS, such as warning strategy selection, system tailor. Using our OEP-LDWS, we assess our LDWS with three different warning strategies, and reach the conclusion that the PTLC is the best under the low false warning criterion.","PeriodicalId":128006,"journal":{"name":"2012 International Conference on Machine Learning and Cybernetics","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128257909","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 : 2012-07-15DOI: 10.1109/ICMLC.2012.6359546
W. Thanadechteemapat, L. Fung
Web Content Extraction technique is proposed in this paper. The technique is able to work with both single and multiple pages based on heuristic rules. An Extracted Content Matching (ECM) technique is proposed in the multiple page extraction to identify the noises among the extracted results. Some features in this technique are also introduced in order to reduce processing time such as use of XPath, file compression, and parallel processing. Assessment of the performance is based on precision, recall and F-measure by using the length of extracted content. Initial results by comparing results from the proposed approach to extraction by manual process are good.
{"title":"Improving Webpage Content Extraction by extending a novel single page extraction approach: A case study with Thai websites","authors":"W. Thanadechteemapat, L. Fung","doi":"10.1109/ICMLC.2012.6359546","DOIUrl":"https://doi.org/10.1109/ICMLC.2012.6359546","url":null,"abstract":"Web Content Extraction technique is proposed in this paper. The technique is able to work with both single and multiple pages based on heuristic rules. An Extracted Content Matching (ECM) technique is proposed in the multiple page extraction to identify the noises among the extracted results. Some features in this technique are also introduced in order to reduce processing time such as use of XPath, file compression, and parallel processing. Assessment of the performance is based on precision, recall and F-measure by using the length of extracted content. Initial results by comparing results from the proposed approach to extraction by manual process are good.","PeriodicalId":128006,"journal":{"name":"2012 International Conference on Machine Learning and Cybernetics","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127138916","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 : 2012-07-15DOI: 10.1109/ICMLC.2012.6359567
Zhiwen Yu, Le Li, Daxing Wang, J. You, Guoqiang Han, Hantao Chen
Clustering ensemble is a momentous technique in machine learning and contribute much to the applications in many areas. General clustering ensemble methods pay more attention to predicting cluster labels than structures of clusters. In fact, learning cluster structures implicates sufficient information to rebuild the dataset and is competent for being the replacement of redundant predicted cluster labels. In this paper, we introduce the fuzzy theory into the structure framework and propose a newfangled double fuzzy c-means structure ensemble framework, named as FCM2SE. FCM2SE makes use of the cluster structure information instead of predicted labels to gain a representative ensemble structure. We also design two novel labeling criteria to distribute the samples to the corresponding clusters. The empirical results on synthetic datasets and UCI machine learning datasets demonstrate the effectiveness of the proposed method.
{"title":"Structure ensemble based on fuzzy c-means","authors":"Zhiwen Yu, Le Li, Daxing Wang, J. You, Guoqiang Han, Hantao Chen","doi":"10.1109/ICMLC.2012.6359567","DOIUrl":"https://doi.org/10.1109/ICMLC.2012.6359567","url":null,"abstract":"Clustering ensemble is a momentous technique in machine learning and contribute much to the applications in many areas. General clustering ensemble methods pay more attention to predicting cluster labels than structures of clusters. In fact, learning cluster structures implicates sufficient information to rebuild the dataset and is competent for being the replacement of redundant predicted cluster labels. In this paper, we introduce the fuzzy theory into the structure framework and propose a newfangled double fuzzy c-means structure ensemble framework, named as FCM2SE. FCM2SE makes use of the cluster structure information instead of predicted labels to gain a representative ensemble structure. We also design two novel labeling criteria to distribute the samples to the corresponding clusters. The empirical results on synthetic datasets and UCI machine learning datasets demonstrate the effectiveness of the proposed method.","PeriodicalId":128006,"journal":{"name":"2012 International Conference on Machine Learning and Cybernetics","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129918388","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 : 2012-07-15DOI: 10.1109/ICMLC.2012.6359668
Fei-Hu Hsieh, Hen-Kung Wang, Po-Lun Chang, H. Wu
This paper proposes related research of the nonlinear dynamic behaviors in voltage-mode controlled single-phase half-bridge inverters. The system exhibits nonlinear dynamic behaviors from period-l operation through period-doubling bifurcations to chaos state using the proportional gain changeable, variable DC input voltage and load resistance. First, a mathematical model of single-phase half-bridge inverter is derived. Then, SIMULINK software tools are used to construct models, simulate results and verify the nonlinear dynamic behaviors in this inverter.
{"title":"Study on dynamic phenomena in voltage-mode controlled single-phase half-bridge inverters","authors":"Fei-Hu Hsieh, Hen-Kung Wang, Po-Lun Chang, H. Wu","doi":"10.1109/ICMLC.2012.6359668","DOIUrl":"https://doi.org/10.1109/ICMLC.2012.6359668","url":null,"abstract":"This paper proposes related research of the nonlinear dynamic behaviors in voltage-mode controlled single-phase half-bridge inverters. The system exhibits nonlinear dynamic behaviors from period-l operation through period-doubling bifurcations to chaos state using the proportional gain changeable, variable DC input voltage and load resistance. First, a mathematical model of single-phase half-bridge inverter is derived. Then, SIMULINK software tools are used to construct models, simulate results and verify the nonlinear dynamic behaviors in this inverter.","PeriodicalId":128006,"journal":{"name":"2012 International Conference on Machine Learning and Cybernetics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129924037","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：