Chris Seiffert, T. Khoshgoftaar, J. V. Hulse, Amri Napolitano
Two common challenges data mining and machine learning practitioners face in many application domains are unequal classification costs and class imbalance. Most traditional data mining techniques attempt to maximize overall accuracy rather than minimize cost. When data is imbalanced, such techniques result in models that highly favor the over represented class, the class which typically carries a lower cost of misclassification. Two techniques that have been used to address both of these issues are cost sensitive learning and data sampling. In this work, we investigate the performance of two cost sensitive learning techniques and four data sampling techniques for minimizing classification costs when data is imbalanced. We present a comprehensive suite of experiments, utilizing 15 datasets with 10 cost ratios, which have been carefully designed to ensure conclusive, significant and reliable results.
{"title":"A Comparative Study of Data Sampling and Cost Sensitive Learning","authors":"Chris Seiffert, T. Khoshgoftaar, J. V. Hulse, Amri Napolitano","doi":"10.1109/ICDMW.2008.119","DOIUrl":"https://doi.org/10.1109/ICDMW.2008.119","url":null,"abstract":"Two common challenges data mining and machine learning practitioners face in many application domains are unequal classification costs and class imbalance. Most traditional data mining techniques attempt to maximize overall accuracy rather than minimize cost. When data is imbalanced, such techniques result in models that highly favor the over represented class, the class which typically carries a lower cost of misclassification. Two techniques that have been used to address both of these issues are cost sensitive learning and data sampling. In this work, we investigate the performance of two cost sensitive learning techniques and four data sampling techniques for minimizing classification costs when data is imbalanced. We present a comprehensive suite of experiments, utilizing 15 datasets with 10 cost ratios, which have been carefully designed to ensure conclusive, significant and reliable results.","PeriodicalId":175955,"journal":{"name":"2008 IEEE International Conference on Data Mining Workshops","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123067592","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}
Advances in computing and communication has resulted in very large scale distributed environments in recent years. They are capable of storing large volumes of data and often have multiple compute nodes. However, the inherent heterogeneity of data components, the dynamic nature of distributed systems, the need for information synchronization and data fusion over a network and security and access control issues makes the problem of resource management and monitoring a tremendous challenge. In particular, centralized algorithms for management of resources and data may not be sufficient to manage complex distributed systems. In this paper, we present a distributed algorithm for resource and data management which builds on the traditional simplex algorithm used for solving linear optimization problems. Our distributed algorithm is an exact one meaning its results are identical if run in a centralized setting. We provide extensive analytical results and experiments on simulated data to demonstrate the performance of our algorithm.
{"title":"Distributed Linear Programming and Resource Management for Data Mining in Distributed Environments","authors":"Haimonti Dutta, H. Kargupta","doi":"10.1109/ICDMW.2008.137","DOIUrl":"https://doi.org/10.1109/ICDMW.2008.137","url":null,"abstract":"Advances in computing and communication has resulted in very large scale distributed environments in recent years. They are capable of storing large volumes of data and often have multiple compute nodes. However, the inherent heterogeneity of data components, the dynamic nature of distributed systems, the need for information synchronization and data fusion over a network and security and access control issues makes the problem of resource management and monitoring a tremendous challenge. In particular, centralized algorithms for management of resources and data may not be sufficient to manage complex distributed systems. In this paper, we present a distributed algorithm for resource and data management which builds on the traditional simplex algorithm used for solving linear optimization problems. Our distributed algorithm is an exact one meaning its results are identical if run in a centralized setting. We provide extensive analytical results and experiments on simulated data to demonstrate the performance of our algorithm.","PeriodicalId":175955,"journal":{"name":"2008 IEEE International Conference on Data Mining Workshops","volume":"118 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125098632","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}
Relations of logical calculi of association rules to measures of interestingness of association rules are studied. Logical calculi of association rules, 4ft-quantifiers and important classes of association rules are briefly introduced. New 4ft-quantifiers and association rules are defined by applications of suitable thresholds to several known measures of interestingness. It is proved that some of new 4ft-quantifiers constitute rules that belong to known classes of rules. It is shown that new interesting classes of rules can be defined on the basis of additional new 4ft-quantifiers. Some additional results concerning new classes of rules are proved. Open problems are introduced.
{"title":"Remarks to Logical Aspects of Measures of Interestingness of Association Rules","authors":"J. Rauch","doi":"10.1109/ICDMW.2008.45","DOIUrl":"https://doi.org/10.1109/ICDMW.2008.45","url":null,"abstract":"Relations of logical calculi of association rules to measures of interestingness of association rules are studied. Logical calculi of association rules, 4ft-quantifiers and important classes of association rules are briefly introduced. New 4ft-quantifiers and association rules are defined by applications of suitable thresholds to several known measures of interestingness. It is proved that some of new 4ft-quantifiers constitute rules that belong to known classes of rules. It is shown that new interesting classes of rules can be defined on the basis of additional new 4ft-quantifiers. Some additional results concerning new classes of rules are proved. Open problems are introduced.","PeriodicalId":175955,"journal":{"name":"2008 IEEE International Conference on Data Mining Workshops","volume":"107 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126073108","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 presents G-REX, a versatile data mining framework based on genetic programming. What differs G-REX from other GP frameworks is that it doesn't strive to be a general purpose framework. This allows G-REX to include more functionality specific to data mining like preprocessing, evaluation- and optimization methods, but also a multitude of predefined classification and regression models. Examples of predefined models are decision trees, decision lists, k-NN with attribute weights, hybrid kNN-rules, fuzzy-rules and several different regression models. The main strength is, however, the flexibility, making it easy to modify, extend and combine all of the predefined functionality. G-REX is, in addition, available in a special Weka package adding useful evolutionary functionality to the standard data mining tool Weka.
{"title":"G-REX: A Versatile Framework for Evolutionary Data Mining","authors":"Rikard König, U. Johansson, L. Niklasson","doi":"10.1109/ICDMW.2008.117","DOIUrl":"https://doi.org/10.1109/ICDMW.2008.117","url":null,"abstract":"This paper presents G-REX, a versatile data mining framework based on genetic programming. What differs G-REX from other GP frameworks is that it doesn't strive to be a general purpose framework. This allows G-REX to include more functionality specific to data mining like preprocessing, evaluation- and optimization methods, but also a multitude of predefined classification and regression models. Examples of predefined models are decision trees, decision lists, k-NN with attribute weights, hybrid kNN-rules, fuzzy-rules and several different regression models. The main strength is, however, the flexibility, making it easy to modify, extend and combine all of the predefined functionality. G-REX is, in addition, available in a special Weka package adding useful evolutionary functionality to the standard data mining tool Weka.","PeriodicalId":175955,"journal":{"name":"2008 IEEE International Conference on Data Mining Workshops","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123428173","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}
We present a demo of ESTER, a search engine that combines the ease of use, speed and scalability of full-text search with the powerful semantic capabilities of ontologies. ESTER supports full-text queries, ontological queries and combinations of these, yet its interface is as easy as can be: A standard search field with semantic information provided interactively as one types. ESTER works by reducing all queries to two basic operations: prefix search and join, which can be implemented very efficiently in terms of both processing time and index space.We demonstrate the capabilities of ESTER on a combination of the English Wikipedia with the Yago ontology, with response times below 100 milliseconds for most queries, and an index size of about 4 GB. The system can be run both stand-alone and as a Web application.
{"title":"Semantic Full-Text Search with ESTER: Scalable, Easy, Fast","authors":"H. Bast, Fabian M. Suchanek, Ingmar Weber","doi":"10.1109/ICDMW.2008.101","DOIUrl":"https://doi.org/10.1109/ICDMW.2008.101","url":null,"abstract":"We present a demo of ESTER, a search engine that combines the ease of use, speed and scalability of full-text search with the powerful semantic capabilities of ontologies. ESTER supports full-text queries, ontological queries and combinations of these, yet its interface is as easy as can be: A standard search field with semantic information provided interactively as one types. ESTER works by reducing all queries to two basic operations: prefix search and join, which can be implemented very efficiently in terms of both processing time and index space.We demonstrate the capabilities of ESTER on a combination of the English Wikipedia with the Yago ontology, with response times below 100 milliseconds for most queries, and an index size of about 4 GB. The system can be run both stand-alone and as a Web application.","PeriodicalId":175955,"journal":{"name":"2008 IEEE International Conference on Data Mining Workshops","volume":"2020 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122187812","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}
With the emergence of large-volume and high-speed streaming data, the recent techniques for stream mining of CFIpsilas (closed frequent itemsets) will become inefficient. When concept drift occurs at a slow rate in high speed data streams, the rate of change of information across different sliding windows will be negligible. So, the user wonpsilat be devoid of change in information if we slide window by multiple transactions at a time. Therefore, we propose a novel approach for mining CFIpsilas cumulatively by making sliding width(ges1) over high speed data streams. However, it is nontrivial to mine CFIpsilas cumulatively over stream, because such growth may lead to the generation of exponential number of candidates for closure checking. In this study, we develop an efficient algorithm, stream-close, for mining CFIpsilas over stream by exploring some interesting properties. Our performance study reveals that stream-close achieves good scalability and has promising results.
随着大容量、高速流数据的出现，现有的封闭频繁项集(CFIpsilas, closed frequency itemset)流挖掘技术将变得低效。当概念漂移在高速数据流中缓慢发生时，信息在不同滑动窗口之间的变化率可以忽略不计。因此，如果我们一次滑动多个事务窗口，用户将无法获得信息更改。因此，我们提出了一种通过在高速数据流上设置滑动宽度(ges1)来累积挖掘CFIpsilas的新方法。然而，在数据流中累积挖掘CFIpsilas是很重要的，因为这种增长可能导致生成指数级的闭包检查候选数据。在这项研究中，我们通过探索一些有趣的性质，开发了一种高效的算法，流关闭，用于挖掘流上的CFIpsilas。我们的性能研究表明，stream-close具有良好的可扩展性和良好的效果。
{"title":"Stream-Close: Fast Mining of Closed Frequent Itemsets in High Speed Data Streams","authors":"Ranganath B. N., M. Murty","doi":"10.1109/ICDMW.2008.51","DOIUrl":"https://doi.org/10.1109/ICDMW.2008.51","url":null,"abstract":"With the emergence of large-volume and high-speed streaming data, the recent techniques for stream mining of CFIpsilas (closed frequent itemsets) will become inefficient. When concept drift occurs at a slow rate in high speed data streams, the rate of change of information across different sliding windows will be negligible. So, the user wonpsilat be devoid of change in information if we slide window by multiple transactions at a time. Therefore, we propose a novel approach for mining CFIpsilas cumulatively by making sliding width(ges1) over high speed data streams. However, it is nontrivial to mine CFIpsilas cumulatively over stream, because such growth may lead to the generation of exponential number of candidates for closure checking. In this study, we develop an efficient algorithm, stream-close, for mining CFIpsilas over stream by exploring some interesting properties. Our performance study reveals that stream-close achieves good scalability and has promising results.","PeriodicalId":175955,"journal":{"name":"2008 IEEE International Conference on Data Mining Workshops","volume":"102 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116719669","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}
We propose a dynamic graph-based relational mining approach using graph-rewriting rules to learns patterns in networks that structurally change over time. A dynamic graph containing a sequence of graphs over time represents dynamic properties as well as structural properties of the network. Our approach discovers graph-rewriting rules, which describe the structural transformations between two sequential graphs over time, and also learns description rules that generalize over the discovered graph-rewriting rules. The discovered graph-rewriting rules show how networks change over time, and the description rules in the graph-rewriting rules show temporal patterns in the structural changes. We apply our approach to biological networks to understand how the biosystems change over time. Our compression-based discovery of the description rules is compared with the frequent subgraph mining approach using several evaluation metrics.
{"title":"Graph-Based Data Mining in Dynamic Networks: Empirical Comparison of Compression-Based and Frequency-Based Subgraph Mining","authors":"C. You, L. Holder, D. Cook","doi":"10.1109/ICDMW.2008.68","DOIUrl":"https://doi.org/10.1109/ICDMW.2008.68","url":null,"abstract":"We propose a dynamic graph-based relational mining approach using graph-rewriting rules to learns patterns in networks that structurally change over time. A dynamic graph containing a sequence of graphs over time represents dynamic properties as well as structural properties of the network. Our approach discovers graph-rewriting rules, which describe the structural transformations between two sequential graphs over time, and also learns description rules that generalize over the discovered graph-rewriting rules. The discovered graph-rewriting rules show how networks change over time, and the description rules in the graph-rewriting rules show temporal patterns in the structural changes. We apply our approach to biological networks to understand how the biosystems change over time. Our compression-based discovery of the description rules is compared with the frequent subgraph mining approach using several evaluation metrics.","PeriodicalId":175955,"journal":{"name":"2008 IEEE International Conference on Data Mining Workshops","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128209389","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 focuses on developing classification algorithms for problems in which there is a need to predict the class based on multiple observations (examples) of the same phenomenon (class). These problems give rise to a new classification problem, referred to as set classification, that requires the prediction of a set of instances given the prior knowledge that all the instances of the set belong to the same unknown class. This problem falls under the general class of problems whose instances have class label dependencies. Four methods for solving the set classification problem are developed and studied. The first is based on a straightforward extension of the traditional classification paradigm whereas the other three are designed to explicitly take into account the known dependencies among the instances of the unlabeled set during learning or classification. A comprehensive experimental evaluation of the various methods and their underlying parameters shows that some of them lead to significant gains in performance.
{"title":"The Set Classification Problem and Solution Methods","authors":"Xia Ning, G. Karypis","doi":"10.1109/ICDMW.2008.113","DOIUrl":"https://doi.org/10.1109/ICDMW.2008.113","url":null,"abstract":"This paper focuses on developing classification algorithms for problems in which there is a need to predict the class based on multiple observations (examples) of the same phenomenon (class). These problems give rise to a new classification problem, referred to as set classification, that requires the prediction of a set of instances given the prior knowledge that all the instances of the set belong to the same unknown class. This problem falls under the general class of problems whose instances have class label dependencies. Four methods for solving the set classification problem are developed and studied. The first is based on a straightforward extension of the traditional classification paradigm whereas the other three are designed to explicitly take into account the known dependencies among the instances of the unlabeled set during learning or classification. A comprehensive experimental evaluation of the various methods and their underlying parameters shows that some of them lead to significant gains in performance.","PeriodicalId":175955,"journal":{"name":"2008 IEEE International Conference on Data Mining Workshops","volume":"74 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128691340","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}
Detecting inconsistent values received in a communication is a challenging problem faced in networked systems. Inconsistent values occur when a message contains incorrect data, even though the syntax is correct and there is no corruption due to transmission errors. In many cases, traditional schemes based on voting protocols or error detection codes cannot be used. An alternative is discovering implicit redundancies, or patterns that model a correct communication, and using these patterns to detect inconsistent values. However, existing techniques do not cover the inputs and sequential patterns needed by this problem. In this paper, we propose a novel technique that considers messages with multiple types and attributes, events involving variables, and a heuristic for reducing redundant information. Experiments show that the discovered redundancies can achieve reasonable error detection coverage in fields where sequential relations exist, without implying in a large number of false alarms or a high latency.
{"title":"Discovering Implicit Redundancies in Network Communications for Detecting Inconsistent Values","authors":"B. Nassu, T. Nanya, Hiroshi Nakamura","doi":"10.1109/ICDMW.2008.15","DOIUrl":"https://doi.org/10.1109/ICDMW.2008.15","url":null,"abstract":"Detecting inconsistent values received in a communication is a challenging problem faced in networked systems. Inconsistent values occur when a message contains incorrect data, even though the syntax is correct and there is no corruption due to transmission errors. In many cases, traditional schemes based on voting protocols or error detection codes cannot be used. An alternative is discovering implicit redundancies, or patterns that model a correct communication, and using these patterns to detect inconsistent values. However, existing techniques do not cover the inputs and sequential patterns needed by this problem. In this paper, we propose a novel technique that considers messages with multiple types and attributes, events involving variables, and a heuristic for reducing redundant information. Experiments show that the discovered redundancies can achieve reasonable error detection coverage in fields where sequential relations exist, without implying in a large number of false alarms or a high latency.","PeriodicalId":175955,"journal":{"name":"2008 IEEE International Conference on Data Mining Workshops","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129905693","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 many practical situations it is not feasible to collect labeled samples for all available classes in a domain. Especially in supervised classification of remotely sensed images it is impossible to collect ground truth information over large geographic regions for all thematic classes. As a result often analysts collect labels for aggregate classes (e.g., Forest, Agriculture, Urban). In this paper we present a novel learning scheme that automatically learns sub-classes (e.g., Hardwood, Conifer) from the user given aggregate classes. We model each aggregate class as finite Gaussian mixture instead of classical assumption of unimodal Gaussian per class. The number of components in each finite Gaussian mixture are automatically estimated. A semi-supervised learning is then used to recognize sub-classes by utilizing very few labeled samples per each sub-class and a large number of unlabeled samples. Experimental results on real remotely sensed image classification showed not only improved accuracy in aggregate class classification but the proposed method also recognized sub-classes accurately.
{"title":"A Semi-supervised Learning Algorithm for Recognizing Sub-classes","authors":"Ranga Raju Vatsavai, S. Shekhar, B. Bhaduri","doi":"10.1109/ICDMW.2008.129","DOIUrl":"https://doi.org/10.1109/ICDMW.2008.129","url":null,"abstract":"In many practical situations it is not feasible to collect labeled samples for all available classes in a domain. Especially in supervised classification of remotely sensed images it is impossible to collect ground truth information over large geographic regions for all thematic classes. As a result often analysts collect labels for aggregate classes (e.g., Forest, Agriculture, Urban). In this paper we present a novel learning scheme that automatically learns sub-classes (e.g., Hardwood, Conifer) from the user given aggregate classes. We model each aggregate class as finite Gaussian mixture instead of classical assumption of unimodal Gaussian per class. The number of components in each finite Gaussian mixture are automatically estimated. A semi-supervised learning is then used to recognize sub-classes by utilizing very few labeled samples per each sub-class and a large number of unlabeled samples. Experimental results on real remotely sensed image classification showed not only improved accuracy in aggregate class classification but the proposed method also recognized sub-classes accurately.","PeriodicalId":175955,"journal":{"name":"2008 IEEE International Conference on Data Mining Workshops","volume":"112 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123465053","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：