Pub Date : 2014-10-20DOI: 10.1109/UKCI.2014.6930181
Donglin Cao, Dazhen Lin, Jiansong Yu
Because of high computational complexity, a long distance spatial-context based automatic image annotation is hard to achieve. Some state of art approaches in image processing, such as 2D-HMM, only considering short distance spatial-context (two neighbors) to reduce the computational complexity. However, these approaches cannot describe long distance semantic spatial-context in image. Therefore, in this paper, we propose a two-step Long Distance Spatial-context Model (LDSM) to solve that problem. First, because of high computational complexity in 2D spatial-context, we transform a 2D spatial-context into a 1D sequence-context. Second, we use conditional random fields to model the 1D sequence-context. Our experiments show that LDSM models the semantic relation between annotated object and background, and experiment results outperform the classical automatic image annotation approach (SVM).
{"title":"Automatic image annotation with long distance spatial-context","authors":"Donglin Cao, Dazhen Lin, Jiansong Yu","doi":"10.1109/UKCI.2014.6930181","DOIUrl":"https://doi.org/10.1109/UKCI.2014.6930181","url":null,"abstract":"Because of high computational complexity, a long distance spatial-context based automatic image annotation is hard to achieve. Some state of art approaches in image processing, such as 2D-HMM, only considering short distance spatial-context (two neighbors) to reduce the computational complexity. However, these approaches cannot describe long distance semantic spatial-context in image. Therefore, in this paper, we propose a two-step Long Distance Spatial-context Model (LDSM) to solve that problem. First, because of high computational complexity in 2D spatial-context, we transform a 2D spatial-context into a 1D sequence-context. Second, we use conditional random fields to model the 1D sequence-context. Our experiments show that LDSM models the semantic relation between annotated object and background, and experiment results outperform the classical automatic image annotation approach (SVM).","PeriodicalId":315044,"journal":{"name":"2014 14th UK Workshop on Computational Intelligence (UKCI)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116970809","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 : 2014-10-20DOI: 10.1109/UKCI.2014.6930157
Richard Hyde, P. Angelov
A new, data density based approach to clustering is presented which automatically determines the number of clusters. By using RDE for each data sample the number of calculations is significantly reduced in offline mode and, further, the method is suitable for online use. The clusters allow a different diameter per feature/dimension creating hyper-ellipsoid clusters which are axis-orthogonal. This results in a greater differentiation between clusters where the clusters are highly asymmetrical. We illustrate this with 3 standard data sets, 1 artificial dataset and a large real dataset to demonstrate comparable results to Subtractive, Hierarchical, K-Means, ELM and DBScan clustering techniques. Unlike subtractive clustering we do not iteratively calculate P however. Unlike hierarchical we do not need O(N2) distances to be calculated and a cut-off threshold to be defined. Unlike k-means we do not need to predefine the number of clusters. Using the RDE equations to calculate the densities the algorithm is efficient, and requires no iteration to approach the optimal result. We compare the proposed algorithm to k-means, subtractive, hierarchical, ELM and DBScan clustering with respect to several criteria. The results demonstrate the validity of the proposed approach.
{"title":"Data density based clustering","authors":"Richard Hyde, P. Angelov","doi":"10.1109/UKCI.2014.6930157","DOIUrl":"https://doi.org/10.1109/UKCI.2014.6930157","url":null,"abstract":"A new, data density based approach to clustering is presented which automatically determines the number of clusters. By using RDE for each data sample the number of calculations is significantly reduced in offline mode and, further, the method is suitable for online use. The clusters allow a different diameter per feature/dimension creating hyper-ellipsoid clusters which are axis-orthogonal. This results in a greater differentiation between clusters where the clusters are highly asymmetrical. We illustrate this with 3 standard data sets, 1 artificial dataset and a large real dataset to demonstrate comparable results to Subtractive, Hierarchical, K-Means, ELM and DBScan clustering techniques. Unlike subtractive clustering we do not iteratively calculate P however. Unlike hierarchical we do not need O(N2) distances to be calculated and a cut-off threshold to be defined. Unlike k-means we do not need to predefine the number of clusters. Using the RDE equations to calculate the densities the algorithm is efficient, and requires no iteration to approach the optimal result. We compare the proposed algorithm to k-means, subtractive, hierarchical, ELM and DBScan clustering with respect to several criteria. The results demonstrate the validity of the proposed approach.","PeriodicalId":315044,"journal":{"name":"2014 14th UK Workshop on Computational Intelligence (UKCI)","volume":"62 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126192029","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 : 2014-10-20DOI: 10.1109/UKCI.2014.6930189
Hamzah S. AlZu'bi, Simon Gerrard-Longworth, W. Al-Nuaimy, J. Goulermas, S. Preece
Human activity recognition is an area of growing interest facilitated by the current revolution in body-worn sensors. Activity recognition allows applications to construct activity profiles for each subject which could be used effectively for healthcare and safety applications. Automated human activity recognition systems face several challenges such as number of sensors, sensor precision, gait style differences, and others. This work proposes a machine learning system to automatically recognise human activities based on a single body-worn accelerometer. The in-house collected dataset contains 3D acceleration of 50 subjects performing 10 different activities. The dataset was produced to ensure robustness and prevent subject-biased results. The feature vector is derived from simple statistical features. The proposed method benefits from RGB-to-YIQ colour space transform as kernel to transform the feature vector into more discriminable features. The classification technique is based on an adaptive boosting ensemble classifier. The proposed system shows consistent classification performance up to 95% accuracy among the 50 subjects.
{"title":"Human activity classification using a single accelerometer","authors":"Hamzah S. AlZu'bi, Simon Gerrard-Longworth, W. Al-Nuaimy, J. Goulermas, S. Preece","doi":"10.1109/UKCI.2014.6930189","DOIUrl":"https://doi.org/10.1109/UKCI.2014.6930189","url":null,"abstract":"Human activity recognition is an area of growing interest facilitated by the current revolution in body-worn sensors. Activity recognition allows applications to construct activity profiles for each subject which could be used effectively for healthcare and safety applications. Automated human activity recognition systems face several challenges such as number of sensors, sensor precision, gait style differences, and others. This work proposes a machine learning system to automatically recognise human activities based on a single body-worn accelerometer. The in-house collected dataset contains 3D acceleration of 50 subjects performing 10 different activities. The dataset was produced to ensure robustness and prevent subject-biased results. The feature vector is derived from simple statistical features. The proposed method benefits from RGB-to-YIQ colour space transform as kernel to transform the feature vector into more discriminable features. The classification technique is based on an adaptive boosting ensemble classifier. The proposed system shows consistent classification performance up to 95% accuracy among the 50 subjects.","PeriodicalId":315044,"journal":{"name":"2014 14th UK Workshop on Computational Intelligence (UKCI)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123625097","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 : 2014-10-20DOI: 10.1109/UKCI.2014.6930178
Luke Vella Critien, Arjab Singh Khuman, Jenny Carter, S. Ahmadi
The aim of this study is to design a rule based expert system which provides doctors with a better tool to be able to manage chronic heart failure in adults. The suggested system is intended to help in diagnosing heart failure at the earliest permissible stage and subsequently suggest the best treatment for the particular case. The designed system has two facets, one related to the diagnosis and the other related to the treatment of chronic heart failure. The former part is based on the latest chronic heart failure guidelines issued by the National Health Services (NHS) - National Institute for Health and Clinical Excellence (NICE) in August 2010. The treatment for chronic heart failure is based on the latest version of British National Formulary (BNF). This rule based system is not intended to replace the specialist but it may be used to provide assurance that all diagnostic criteria have been followed and hence the best possible treatment is given. The system is implemented using the CLIPS language which is a powerful forward-chaining rule based system.
{"title":"A rule based system for diagnosing and treating chronic heart failure","authors":"Luke Vella Critien, Arjab Singh Khuman, Jenny Carter, S. Ahmadi","doi":"10.1109/UKCI.2014.6930178","DOIUrl":"https://doi.org/10.1109/UKCI.2014.6930178","url":null,"abstract":"The aim of this study is to design a rule based expert system which provides doctors with a better tool to be able to manage chronic heart failure in adults. The suggested system is intended to help in diagnosing heart failure at the earliest permissible stage and subsequently suggest the best treatment for the particular case. The designed system has two facets, one related to the diagnosis and the other related to the treatment of chronic heart failure. The former part is based on the latest chronic heart failure guidelines issued by the National Health Services (NHS) - National Institute for Health and Clinical Excellence (NICE) in August 2010. The treatment for chronic heart failure is based on the latest version of British National Formulary (BNF). This rule based system is not intended to replace the specialist but it may be used to provide assurance that all diagnostic criteria have been followed and hence the best possible treatment is given. The system is implemented using the CLIPS language which is a powerful forward-chaining rule based system.","PeriodicalId":315044,"journal":{"name":"2014 14th UK Workshop on Computational Intelligence (UKCI)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123444850","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 : 2014-10-20DOI: 10.1109/UKCI.2014.6930165
Yousef Alqasrawi, D. Neagu
The problem of image annotation has gained increasing attention from many researchers in computer vision. Few works have addressed the use of bag of visual words for scene annotation at region level. The aim of this paper is to study the relationship between the distribution of local semantic concepts and local keypoints located in image regions labelled with these semantic concepts. Based on this study, we investigate whether bag of visual words model can be used to efficiently represent the content of natural scene image regions, so images can be annotated with local semantic concepts. Also, this paper presents local from global approach which study the influence of using visual vocabularies generated from general scene categories to build bag of visual words at region level. Extensive experiments are conducted over a natural scene dataset with six categories. The reported results have shown the plausibility of using the BOW model to represent the semantic information of image regions.
{"title":"Investigating the relationship between the distribution of local semantic concepts and local keypoints for image annotation","authors":"Yousef Alqasrawi, D. Neagu","doi":"10.1109/UKCI.2014.6930165","DOIUrl":"https://doi.org/10.1109/UKCI.2014.6930165","url":null,"abstract":"The problem of image annotation has gained increasing attention from many researchers in computer vision. Few works have addressed the use of bag of visual words for scene annotation at region level. The aim of this paper is to study the relationship between the distribution of local semantic concepts and local keypoints located in image regions labelled with these semantic concepts. Based on this study, we investigate whether bag of visual words model can be used to efficiently represent the content of natural scene image regions, so images can be annotated with local semantic concepts. Also, this paper presents local from global approach which study the influence of using visual vocabularies generated from general scene categories to build bag of visual words at region level. Extensive experiments are conducted over a natural scene dataset with six categories. The reported results have shown the plausibility of using the BOW model to represent the semantic information of image regions.","PeriodicalId":315044,"journal":{"name":"2014 14th UK Workshop on Computational Intelligence (UKCI)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125776741","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 : 2014-10-20DOI: 10.1109/UKCI.2014.6930159
Lifei Chen, G. Guo, Shengrui Wang, Xiangzeng Kong
Kernel-based methods have become popular in machine learning; however, they are typically designed for numeric data. These methods are established in vector spaces, which are undefined for categorical data. In this paper, we propose a new kind of kernel trick, showing that mapping of categorical samples into kernel spaces can be alternatively described as assigning a kernel-based weight to each categorical attribute of the input space, so that common distance measures can be employed. A data-driven approach is then proposed to kernel bandwidth selection by optimizing feature weights. We also make use of the kernel-based distance measure to effectively extend nearest-neighbor classification to classify categorical data. Experimental results on real-world data sets show the outstanding performance of this approach compared to that obtained in the original input space.
{"title":"Kernel learning method for distance-based classification of categorical data","authors":"Lifei Chen, G. Guo, Shengrui Wang, Xiangzeng Kong","doi":"10.1109/UKCI.2014.6930159","DOIUrl":"https://doi.org/10.1109/UKCI.2014.6930159","url":null,"abstract":"Kernel-based methods have become popular in machine learning; however, they are typically designed for numeric data. These methods are established in vector spaces, which are undefined for categorical data. In this paper, we propose a new kind of kernel trick, showing that mapping of categorical samples into kernel spaces can be alternatively described as assigning a kernel-based weight to each categorical attribute of the input space, so that common distance measures can be employed. A data-driven approach is then proposed to kernel bandwidth selection by optimizing feature weights. We also make use of the kernel-based distance measure to effectively extend nearest-neighbor classification to classify categorical data. Experimental results on real-world data sets show the outstanding performance of this approach compared to that obtained in the original input space.","PeriodicalId":315044,"journal":{"name":"2014 14th UK Workshop on Computational Intelligence (UKCI)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129740265","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 : 2014-10-20DOI: 10.1109/UKCI.2014.6930174
Jayne Eaton, Shengxiang Yang
Efficient rescheduling after a perturbation is an important concern of the railway industry. Extreme delays can result in large fines for the train company as well as dissatisfied customers. The problem is exacerbated by the fact that it is a dynamic one; more timetabled trains may be arriving as the perturbed trains are waiting to be rescheduled. The new trains may have different priorities to the existing trains and thus the rescheduling problem is a dynamic one that changes over time. The aim of this research is to apply a population-based ant colony optimisation algorithm to address this dynamic railway junction rescheduling problem using a simulator modelled on a real-world junction in the UK railway network. The results are promising: the algorithm performs well, particularly when the dynamic changes are of a high magnitude and frequency.
{"title":"Dynamic railway junction rescheduling using population based ant colony optimisation","authors":"Jayne Eaton, Shengxiang Yang","doi":"10.1109/UKCI.2014.6930174","DOIUrl":"https://doi.org/10.1109/UKCI.2014.6930174","url":null,"abstract":"Efficient rescheduling after a perturbation is an important concern of the railway industry. Extreme delays can result in large fines for the train company as well as dissatisfied customers. The problem is exacerbated by the fact that it is a dynamic one; more timetabled trains may be arriving as the perturbed trains are waiting to be rescheduled. The new trains may have different priorities to the existing trains and thus the rescheduling problem is a dynamic one that changes over time. The aim of this research is to apply a population-based ant colony optimisation algorithm to address this dynamic railway junction rescheduling problem using a simulator modelled on a real-world junction in the UK railway network. The results are promising: the algorithm performs well, particularly when the dynamic changes are of a high magnitude and frequency.","PeriodicalId":315044,"journal":{"name":"2014 14th UK Workshop on Computational Intelligence (UKCI)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121404824","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 : 2014-10-20DOI: 10.1109/UKCI.2014.6930164
P. Faragó, G. Oltean, L. Ivanciu
To fully analyze the time response of a complex system, in order to discover its critical operation points, the output waveform (under all conceivable conditions) needs to be generated. Using conventional methods as physical experiments or detailed simulations can be prohibitive from the resources point of view (time, equipment). The challenge is to generate the waveform by its numerous time samples as a function of different operating conditions described by a set of parameters. In this paper, we propose a fast to evaluate, but also accurate model that approximates the waveforms, as a reliable substitute for complex physical experiments or overwhelming system simulations. Our proposed model consists of two stages. In the first stage, a previously trained artificial neural network produces some coefficients standing for “primary” coefficients of a wavelet transform. In the second stage, an inverse wavelet transform generates all the time samples of the expected waveform, using a fusion between the “primary” coefficients and some “secondary” coefficients previously extracted from the nominal waveform in the family. The test results for a number of 100 different combinations of three waveform parameters show that our model is a reliable one, featuring high accuracy and generalization capabilities, as well as high computation speed.
{"title":"Neural networks and wavelet transform in waveform approximation","authors":"P. Faragó, G. Oltean, L. Ivanciu","doi":"10.1109/UKCI.2014.6930164","DOIUrl":"https://doi.org/10.1109/UKCI.2014.6930164","url":null,"abstract":"To fully analyze the time response of a complex system, in order to discover its critical operation points, the output waveform (under all conceivable conditions) needs to be generated. Using conventional methods as physical experiments or detailed simulations can be prohibitive from the resources point of view (time, equipment). The challenge is to generate the waveform by its numerous time samples as a function of different operating conditions described by a set of parameters. In this paper, we propose a fast to evaluate, but also accurate model that approximates the waveforms, as a reliable substitute for complex physical experiments or overwhelming system simulations. Our proposed model consists of two stages. In the first stage, a previously trained artificial neural network produces some coefficients standing for “primary” coefficients of a wavelet transform. In the second stage, an inverse wavelet transform generates all the time samples of the expected waveform, using a fusion between the “primary” coefficients and some “secondary” coefficients previously extracted from the nominal waveform in the family. The test results for a number of 100 different combinations of three waveform parameters show that our model is a reliable one, featuring high accuracy and generalization capabilities, as well as high computation speed.","PeriodicalId":315044,"journal":{"name":"2014 14th UK Workshop on Computational Intelligence (UKCI)","volume":"106 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124818151","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 : 2014-10-20DOI: 10.1109/UKCI.2014.6930160
O. Ibrahim, Dario Landa Silva
This paper introduces a new weighting scheme in information retrieval. It also proposes using the document centroid as a threshold for normalizing documents in a document collection. Document centroid normalization helps to achieve more effective information retrieval as it enables good discrimination between documents. In the context of a machine learning application, namely unsupervised document indexing and retrieval, we compared the effectiveness of the proposed weighting scheme to the `Term Frequency - Inverse Document Frequency' or TF-IDF, which is commonly used and considered as one of the best existing weighting schemes. The paper shows how the document centroid is used to remove less significant weights from documents and how this helps to achieve better retrieval effectiveness. Most of the existing weighting schemes in information retrieval research assume that the whole document collection is static. The results presented in this paper show that the proposed weighting scheme can produce higher retrieval effectiveness compared with the TF-IDF weighting scheme, in both static and dynamic document collections. The results also show the variation in information retrieval effectiveness that is achieved for static and dynamic document collections by using a specific weighting scheme. This type of comparison has not been presented in the literature before.
{"title":"A new weighting scheme and discriminative approach for information retrieval in static and dynamic document collections","authors":"O. Ibrahim, Dario Landa Silva","doi":"10.1109/UKCI.2014.6930160","DOIUrl":"https://doi.org/10.1109/UKCI.2014.6930160","url":null,"abstract":"This paper introduces a new weighting scheme in information retrieval. It also proposes using the document centroid as a threshold for normalizing documents in a document collection. Document centroid normalization helps to achieve more effective information retrieval as it enables good discrimination between documents. In the context of a machine learning application, namely unsupervised document indexing and retrieval, we compared the effectiveness of the proposed weighting scheme to the `Term Frequency - Inverse Document Frequency' or TF-IDF, which is commonly used and considered as one of the best existing weighting schemes. The paper shows how the document centroid is used to remove less significant weights from documents and how this helps to achieve better retrieval effectiveness. Most of the existing weighting schemes in information retrieval research assume that the whole document collection is static. The results presented in this paper show that the proposed weighting scheme can produce higher retrieval effectiveness compared with the TF-IDF weighting scheme, in both static and dynamic document collections. The results also show the variation in information retrieval effectiveness that is achieved for static and dynamic document collections by using a specific weighting scheme. This type of comparison has not been presented in the literature before.","PeriodicalId":315044,"journal":{"name":"2014 14th UK Workshop on Computational Intelligence (UKCI)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124032086","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 : 2014-10-20DOI: 10.1109/UKCI.2014.6930182
Ray-Ming Chen
Principal Component Analysis (PCA) is an important technique in finding uncorrelated variables. It is applied in many fields: machine learning, pattern recognition, data mining, compression, ..., etc. In this paper, we introduce this technique into approximation reasoning. Before the introduction, we construct a theoretical framework of such approximation first. This approximation is based on reasoning of incomplete information in which there exists no algorithm such that the intersection between arbitrary target sets and partitioned clusters is decidable, while there exist some algorithms for the decidability of the subset operation between them. Then, under this framework, we utilize PCA to implement such approximation reasoning. PCA is mainly applied to partitioning a universe repeatedly until all the partitioned sets are singular or indecomposable. Then we collect all the partitioned clusters as the granular knowledge and then use this knowledge to approximate the target set.
{"title":"PCA-based algorithmic approximation of crisp target sets","authors":"Ray-Ming Chen","doi":"10.1109/UKCI.2014.6930182","DOIUrl":"https://doi.org/10.1109/UKCI.2014.6930182","url":null,"abstract":"Principal Component Analysis (PCA) is an important technique in finding uncorrelated variables. It is applied in many fields: machine learning, pattern recognition, data mining, compression, ..., etc. In this paper, we introduce this technique into approximation reasoning. Before the introduction, we construct a theoretical framework of such approximation first. This approximation is based on reasoning of incomplete information in which there exists no algorithm such that the intersection between arbitrary target sets and partitioned clusters is decidable, while there exist some algorithms for the decidability of the subset operation between them. Then, under this framework, we utilize PCA to implement such approximation reasoning. PCA is mainly applied to partitioning a universe repeatedly until all the partitioned sets are singular or indecomposable. Then we collect all the partitioned clusters as the granular knowledge and then use this knowledge to approximate the target set.","PeriodicalId":315044,"journal":{"name":"2014 14th UK Workshop on Computational Intelligence (UKCI)","volume":"73 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131058563","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：