In this paper, we present a novel feature selection approach based on an efficient selection of pair wise constraints. This aims at selecting the most coherent constraints extracted from labeled part of data. The relevance of features is then evaluated according to their efficient locality preserving and chosen constraint preserving ability. Finally, experimental results are provided for validating our proposal with respect to other known feature selection methods.
{"title":"Constraint Selection-Based Semi-supervised Feature Selection","authors":"Mohammed Hindawi, Kais Allab, K. Benabdeslem","doi":"10.1109/ICDM.2011.42","DOIUrl":"https://doi.org/10.1109/ICDM.2011.42","url":null,"abstract":"In this paper, we present a novel feature selection approach based on an efficient selection of pair wise constraints. This aims at selecting the most coherent constraints extracted from labeled part of data. The relevance of features is then evaluated according to their efficient locality preserving and chosen constraint preserving ability. Finally, experimental results are provided for validating our proposal with respect to other known feature selection methods.","PeriodicalId":106216,"journal":{"name":"2011 IEEE 11th International Conference on Data Mining","volume":"276 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116070431","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}
Yexi Jiang, Chang-Shing Perng, Tao Li, Rong N. Chang
The promise of cloud computing is to provide computing resources instantly whenever they are needed. The state-of-art virtual machine (VM) provisioning technology can provision a VM in tens of minutes. This latency is unacceptable for jobs that need to scale out during computation. To truly enable on-the-fly scaling, new VM needs to be ready in seconds upon request. In this paper, We present an online temporal data mining system called ASAP, to model and predict the cloud VM demands. ASAP aims to extract high level characteristics from VM provisioning request stream and notify the provisioning system to prepare VMs in advance. For quantification issue, we propose Cloud Prediction Cost to encodes the cost and constraints of the cloud and guide the training of prediction algorithms. Moreover, we utilize a two-level ensemble method to capture the characteristics of the high transient demands time series. Experimental results using historical data from an IBM cloud in operation demonstrate that ASAP significantly improves the cloud service quality and provides possibility for on-the-fly provisioning.
{"title":"ASAP: A Self-Adaptive Prediction System for Instant Cloud Resource Demand Provisioning","authors":"Yexi Jiang, Chang-Shing Perng, Tao Li, Rong N. Chang","doi":"10.1109/ICDM.2011.25","DOIUrl":"https://doi.org/10.1109/ICDM.2011.25","url":null,"abstract":"The promise of cloud computing is to provide computing resources instantly whenever they are needed. The state-of-art virtual machine (VM) provisioning technology can provision a VM in tens of minutes. This latency is unacceptable for jobs that need to scale out during computation. To truly enable on-the-fly scaling, new VM needs to be ready in seconds upon request. In this paper, We present an online temporal data mining system called ASAP, to model and predict the cloud VM demands. ASAP aims to extract high level characteristics from VM provisioning request stream and notify the provisioning system to prepare VMs in advance. For quantification issue, we propose Cloud Prediction Cost to encodes the cost and constraints of the cloud and guide the training of prediction algorithms. Moreover, we utilize a two-level ensemble method to capture the characteristics of the high transient demands time series. Experimental results using historical data from an IBM cloud in operation demonstrate that ASAP significantly improves the cloud service quality and provides possibility for on-the-fly provisioning.","PeriodicalId":106216,"journal":{"name":"2011 IEEE 11th International Conference on Data Mining","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126414909","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}
To understand how social networks evolve over time, graphs representing the networks need to be published periodically or on-demand. The identity of the participants (nodes) must be anonymized to protect the privacy of the individuals and their relationships (edges) to the other members in the social network. We identify a new form of privacy attack, which we name the degree-trail attack. This attack re-identifies the nodes belonging to a target participant from a sequence of published graphs by comparing the degree of the nodes in the published graphs with the degree evolution of a target. The power of this attack is that the adversary can actively influence the degree of the target individual by interacting with the social network. We show that the adversary can succeed with a high probability even if published graphs are anonymized by strongest known privacy preserving techniques in the literature. Moreover, this success does not depend on the distinctiveness of the target nodes nor require the adversary to behave differently from a normal participant. One of our contributions is a formal method to assess the privacy risk of this type of attacks and empirically study the severity on real social network data.
{"title":"Privacy Risk in Graph Stream Publishing for Social Network Data","authors":"Nigel Medforth, Ke Wang","doi":"10.1109/ICDM.2011.120","DOIUrl":"https://doi.org/10.1109/ICDM.2011.120","url":null,"abstract":"To understand how social networks evolve over time, graphs representing the networks need to be published periodically or on-demand. The identity of the participants (nodes) must be anonymized to protect the privacy of the individuals and their relationships (edges) to the other members in the social network. We identify a new form of privacy attack, which we name the degree-trail attack. This attack re-identifies the nodes belonging to a target participant from a sequence of published graphs by comparing the degree of the nodes in the published graphs with the degree evolution of a target. The power of this attack is that the adversary can actively influence the degree of the target individual by interacting with the social network. We show that the adversary can succeed with a high probability even if published graphs are anonymized by strongest known privacy preserving techniques in the literature. Moreover, this success does not depend on the distinctiveness of the target nodes nor require the adversary to behave differently from a normal participant. One of our contributions is a formal method to assess the privacy risk of this type of attacks and empirically study the severity on real social network data.","PeriodicalId":106216,"journal":{"name":"2011 IEEE 11th International Conference on Data Mining","volume":"80 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133872377","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}
Vector Space Model (VSM) is widely used to represent documents and web pages. It is simple and easy to deal computationally, but it also oversimplifies a document into a vector, susceptible to noise, and cannot explicitly represent underlying topics of a document. A matrix representation of document is proposed in this paper: rows represent distinct terms and columns represent cohesive segments. The matrix model views a document as a set of segments, and each segment is a probability distribution over a limited number of latent topics which can be mapped to clustering structures. The latent topic extraction based on the matrix representation of documents is formulated as a constraint optimization problem in which each matrix (i.e., a document) A_i is factorized into a common base determined by non-negative matrices L and R^top, and a non-negative weight matrix M_i such that the sum of reconstruction error on all documents is minimized. Empirical evaluation demonstrates that it is feasible to use the matrix model for document clustering: (1) compared with vector representation, using matrix representation improves clustering quality consistently, and the proposed approach achieves a relative accuracy improvement up to 66% on the studied datasets, and (2) the proposed method outperforms baseline methods such as k-means and NMF, and complements the state-of-the-art methods like LDA and PLSI. Furthermore, the proposed matrix model allows more refined information retrieval at a segment level instead of at a document level, which enables the return of more relevant documents in information retrieval tasks.
{"title":"Document Clustering via Matrix Representation","authors":"Xufei Wang, Jiliang Tang, Huan Liu","doi":"10.1109/ICDM.2011.59","DOIUrl":"https://doi.org/10.1109/ICDM.2011.59","url":null,"abstract":"Vector Space Model (VSM) is widely used to represent documents and web pages. It is simple and easy to deal computationally, but it also oversimplifies a document into a vector, susceptible to noise, and cannot explicitly represent underlying topics of a document. A matrix representation of document is proposed in this paper: rows represent distinct terms and columns represent cohesive segments. The matrix model views a document as a set of segments, and each segment is a probability distribution over a limited number of latent topics which can be mapped to clustering structures. The latent topic extraction based on the matrix representation of documents is formulated as a constraint optimization problem in which each matrix (i.e., a document) A_i is factorized into a common base determined by non-negative matrices L and R^top, and a non-negative weight matrix M_i such that the sum of reconstruction error on all documents is minimized. Empirical evaluation demonstrates that it is feasible to use the matrix model for document clustering: (1) compared with vector representation, using matrix representation improves clustering quality consistently, and the proposed approach achieves a relative accuracy improvement up to 66% on the studied datasets, and (2) the proposed method outperforms baseline methods such as k-means and NMF, and complements the state-of-the-art methods like LDA and PLSI. Furthermore, the proposed matrix model allows more refined information retrieval at a segment level instead of at a document level, which enables the return of more relevant documents in information retrieval tasks.","PeriodicalId":106216,"journal":{"name":"2011 IEEE 11th International Conference on Data Mining","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133274488","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, we propose an efficient and robust algorithm for graph-based transductive classification. After approximating a graph with a spanning tree, we develop a linear-time algorithm to label the tree such that the cut size of the tree is minimized. This significantly improves typical graph-based methods, which either have a cubic time complexity (for a dense graph) or $O(kn^2)$ (for a sparse graph with $k$ denoting the node degree). %In addition to its great scalability on large data, our proposed algorithm demonstrates high robustness and accuracy. In particular, on a graph with 400,000 nodes (in which 10,000 nodes are labeled) and 10,455,545 edges, our algorithm achieves the highest accuracy of $99.6%$ but takes less than $10$ seconds to label all the unlabeled data. Furthermore, our method shows great robustness to the graph construction both theoretically and empirically, this overcomes another big problem of traditional graph-based methods. In addition to its good scalability and robustness, the proposed algorithm demonstrates high accuracy. In particular, on a graph with $400,000$ nodes (in which $10,000$ nodes are labeled) and $10,455,545$ edges, our algorithm achieves the highest accuracy of $99.6%$ but takes less than $10$ seconds to label all the unlabeled data.
{"title":"Fast and Robust Graph-based Transductive Learning via Minimum Tree Cut","authors":"Yanming Zhang, Kaizhu Huang, Cheng-Lin Liu","doi":"10.1109/ICDM.2011.66","DOIUrl":"https://doi.org/10.1109/ICDM.2011.66","url":null,"abstract":"In this paper, we propose an efficient and robust algorithm for graph-based transductive classification. After approximating a graph with a spanning tree, we develop a linear-time algorithm to label the tree such that the cut size of the tree is minimized. This significantly improves typical graph-based methods, which either have a cubic time complexity (for a dense graph) or $O(kn^2)$ (for a sparse graph with $k$ denoting the node degree). %In addition to its great scalability on large data, our proposed algorithm demonstrates high robustness and accuracy. In particular, on a graph with 400,000 nodes (in which 10,000 nodes are labeled) and 10,455,545 edges, our algorithm achieves the highest accuracy of $99.6%$ but takes less than $10$ seconds to label all the unlabeled data. Furthermore, our method shows great robustness to the graph construction both theoretically and empirically, this overcomes another big problem of traditional graph-based methods. In addition to its good scalability and robustness, the proposed algorithm demonstrates high accuracy. In particular, on a graph with $400,000$ nodes (in which $10,000$ nodes are labeled) and $10,455,545$ edges, our algorithm achieves the highest accuracy of $99.6%$ but takes less than $10$ seconds to label all the unlabeled data.","PeriodicalId":106216,"journal":{"name":"2011 IEEE 11th International Conference on Data Mining","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116310771","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}
Multi-instance learning, like other machine learning and data mining tasks, requires distance metrics. Although metric learning methods have been studied for many years, metric learners for multi-instance learning remain almost untouched. In this paper, we propose a framework called Multi-Instance MEtric Learning (MIMEL) to learn an appropriate distance under the multi-instance setting. The distance metric between two bags is defined using the Mahalanobis distance function. The problem is formulated by minimizing the KL divergence between two multivariate Gaussians under the constraints of maximizing the between-class bag distance and minimizing the within-class bag distance. To exploit the mechanism of how instances determine bag labels in multi-instance learning, we design a nonparametric density-estimation-based weighting scheme to assign higher “weights†to the instances that are more likely to be positive in positive bags. The weighting scheme itself has a small workload, which adds little extra computing costs to the proposed framework. Moreover, to further boost the classification accuracy, a kernel version of MIMEL is presented. We evaluate MIMEL, using not only several typical multi-instance tasks, but also two activity recognition datasets. The experimental results demonstrate that MIMEL achieves better classification accuracy than many state-of-the-art distance based algorithms or kernel methods for multi-instance learning.
{"title":"Multi-instance Metric Learning","authors":"Ye Xu, Wei Ping, A. Campbell","doi":"10.1109/ICDM.2011.106","DOIUrl":"https://doi.org/10.1109/ICDM.2011.106","url":null,"abstract":"Multi-instance learning, like other machine learning and data mining tasks, requires distance metrics. Although metric learning methods have been studied for many years, metric learners for multi-instance learning remain almost untouched. In this paper, we propose a framework called Multi-Instance MEtric Learning (MIMEL) to learn an appropriate distance under the multi-instance setting. The distance metric between two bags is defined using the Mahalanobis distance function. The problem is formulated by minimizing the KL divergence between two multivariate Gaussians under the constraints of maximizing the between-class bag distance and minimizing the within-class bag distance. To exploit the mechanism of how instances determine bag labels in multi-instance learning, we design a nonparametric density-estimation-based weighting scheme to assign higher “weights†to the instances that are more likely to be positive in positive bags. The weighting scheme itself has a small workload, which adds little extra computing costs to the proposed framework. Moreover, to further boost the classification accuracy, a kernel version of MIMEL is presented. We evaluate MIMEL, using not only several typical multi-instance tasks, but also two activity recognition datasets. The experimental results demonstrate that MIMEL achieves better classification accuracy than many state-of-the-art distance based algorithms or kernel methods for multi-instance learning.","PeriodicalId":106216,"journal":{"name":"2011 IEEE 11th International Conference on Data Mining","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124860881","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 spectrum of graph has been widely used in graph mining to extract graph topological information. It has also been employed as a characteristic of graph to check the sub graph isomorphism testing since it is an invariant of a graph. However, the spectrum cannot be directly applied to a graph and its sub graph, which is a bottleneck for sub graph isomorphism testing. In this paper, we study the Laplacian spectra between a graph and its sub graph, and propose a method by straightforward adoption of them for sub graph queries. In our proposed method, we first encode every vertex and graph by extracting their Laplacian spectra, and generate a novel two-step filtering conditions. Then, we follow the filtering-and verification framework to conduct sub graph queries. Extensive experiments show that, compared with existing counterpart method, as a graph feature, Laplacian spectra can be used to efficiently improves the efficiency of sub graph queries and thus indicate that it have considerable potential.
{"title":"A Study of Laplacian Spectra of Graph for Subgraph Queries","authors":"Lei Zhu, Qinbao Song","doi":"10.1109/ICDM.2011.17","DOIUrl":"https://doi.org/10.1109/ICDM.2011.17","url":null,"abstract":"The spectrum of graph has been widely used in graph mining to extract graph topological information. It has also been employed as a characteristic of graph to check the sub graph isomorphism testing since it is an invariant of a graph. However, the spectrum cannot be directly applied to a graph and its sub graph, which is a bottleneck for sub graph isomorphism testing. In this paper, we study the Laplacian spectra between a graph and its sub graph, and propose a method by straightforward adoption of them for sub graph queries. In our proposed method, we first encode every vertex and graph by extracting their Laplacian spectra, and generate a novel two-step filtering conditions. Then, we follow the filtering-and verification framework to conduct sub graph queries. Extensive experiments show that, compared with existing counterpart method, as a graph feature, Laplacian spectra can be used to efficiently improves the efficiency of sub graph queries and thus indicate that it have considerable potential.","PeriodicalId":106216,"journal":{"name":"2011 IEEE 11th International Conference on Data Mining","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122064074","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}
Complex systems have been widely studied to characterize their structural behaviors from a topological perspective. High modularity is one of the recurrent features of real-world complex systems. Various graph clustering algorithms have been applied to identifying communities in social networks or modules in biological networks. However, their applicability to real-world systems has been limited because of the massive scale and complex connectivity of the networks. In this study, we exploit a novel information-theoretic model for graph clustering. The entropy-based clustering approach finds locally optimal clusters by growing a random seed in a manner that minimizes graph entropy. We design and analyze modifications that further improve its performance. Assigning priority in seed-selection and seed-growth is well applicable to the scale-free networks characterized by the hub-oriented structure. Computing seed-growth in parallel streams also decomposes an extremely large network efficiently. The experimental results with real biological and social networks show that the entropy-based approach has better performance than competing methods in terms of accuracy and efficiency.
{"title":"Entropy-Based Graph Clustering: Application to Biological and Social Networks","authors":"Edward Casey Kenley, Young-Rae Cho","doi":"10.1109/ICDM.2011.64","DOIUrl":"https://doi.org/10.1109/ICDM.2011.64","url":null,"abstract":"Complex systems have been widely studied to characterize their structural behaviors from a topological perspective. High modularity is one of the recurrent features of real-world complex systems. Various graph clustering algorithms have been applied to identifying communities in social networks or modules in biological networks. However, their applicability to real-world systems has been limited because of the massive scale and complex connectivity of the networks. In this study, we exploit a novel information-theoretic model for graph clustering. The entropy-based clustering approach finds locally optimal clusters by growing a random seed in a manner that minimizes graph entropy. We design and analyze modifications that further improve its performance. Assigning priority in seed-selection and seed-growth is well applicable to the scale-free networks characterized by the hub-oriented structure. Computing seed-growth in parallel streams also decomposes an extremely large network efficiently. The experimental results with real biological and social networks show that the entropy-based approach has better performance than competing methods in terms of accuracy and efficiency.","PeriodicalId":106216,"journal":{"name":"2011 IEEE 11th International Conference on Data Mining","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125268842","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}
Communities are natural structures observed in social networks and are usually characterized as "relatively dense" subsets of nodes. Social networks change over time and so do the underlying community structures. Thus, to truly uncover this structure we must take the temporal aspect of networks into consideration. Previously, we have represented framework for finding dynamic communities using the social cost model and formulated the corresponding optimization problem [33], assuming that partitions of individuals into groups are given in each time step. We have also presented heuristics and approximation algorithms for the problem, with the same assumption [32]. In general, however, dynamic social networks are represented as a sequence of graphs of snapshots of the social network and the assumption that we have partitions of individuals into groups does not hold. In this paper, we extend the social cost model and formulate an optimization problem of finding community structure from the sequence of arbitrary graphs. We propose a semi definite programming formulation and a heuristic rounding scheme. We show, using synthetic data sets, that this method is quite accurate on synthetic data sets and present its results on a real social network.
{"title":"Finding Communities in Dynamic Social Networks","authors":"Chayant Tantipathananandh, T. Berger-Wolf","doi":"10.1109/ICDM.2011.67","DOIUrl":"https://doi.org/10.1109/ICDM.2011.67","url":null,"abstract":"Communities are natural structures observed in social networks and are usually characterized as \"relatively dense\" subsets of nodes. Social networks change over time and so do the underlying community structures. Thus, to truly uncover this structure we must take the temporal aspect of networks into consideration. Previously, we have represented framework for finding dynamic communities using the social cost model and formulated the corresponding optimization problem [33], assuming that partitions of individuals into groups are given in each time step. We have also presented heuristics and approximation algorithms for the problem, with the same assumption [32]. In general, however, dynamic social networks are represented as a sequence of graphs of snapshots of the social network and the assumption that we have partitions of individuals into groups does not hold. In this paper, we extend the social cost model and formulate an optimization problem of finding community structure from the sequence of arbitrary graphs. We propose a semi definite programming formulation and a heuristic rounding scheme. We show, using synthetic data sets, that this method is quite accurate on synthetic data sets and present its results on a real social network.","PeriodicalId":106216,"journal":{"name":"2011 IEEE 11th International Conference on Data Mining","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125409889","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, we present a semi-supervised method to learn a low rank Mahalanobis distance function. Based on an approximation to the projection distance from a manifold, we propose a novel parametric manifold regularizer. In contrast to previous approaches that usually exploit side information only, our proposed method can further take advantages of the intrinsic manifold information from data. In addition, we focus on learning a metric of low rank directly, this is different from traditional approaches that often enforce the l_1 norm on the metric. The resulting configuration is convex with respect to the manifold structure and the distance function, respectively. We solve it with an alternating optimization algorithm, which proves effective to find a satisfactory solution. For efficient implementation, we even present a fast algorithm, in which the manifold structure and the distance function are learned independently without alternating minimization. Experimental results over 12 standard UCI data sets demonstrate the advantages of our method.
{"title":"Low Rank Metric Learning with Manifold Regularization","authors":"G. Zhong, Kaizhu Huang, Cheng-Lin Liu","doi":"10.1109/ICDM.2011.95","DOIUrl":"https://doi.org/10.1109/ICDM.2011.95","url":null,"abstract":"In this paper, we present a semi-supervised method to learn a low rank Mahalanobis distance function. Based on an approximation to the projection distance from a manifold, we propose a novel parametric manifold regularizer. In contrast to previous approaches that usually exploit side information only, our proposed method can further take advantages of the intrinsic manifold information from data. In addition, we focus on learning a metric of low rank directly, this is different from traditional approaches that often enforce the l_1 norm on the metric. The resulting configuration is convex with respect to the manifold structure and the distance function, respectively. We solve it with an alternating optimization algorithm, which proves effective to find a satisfactory solution. For efficient implementation, we even present a fast algorithm, in which the manifold structure and the distance function are learned independently without alternating minimization. Experimental results over 12 standard UCI data sets demonstrate the advantages of our method.","PeriodicalId":106216,"journal":{"name":"2011 IEEE 11th International Conference on Data Mining","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117097729","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：