Sun, 06 Jul 2008 02:32:17 BST CiteULike: sona's library [86 articles] http://www.citeulike.org/user/sona CiteULike.org en-gb Copyright © 2004-2008 citeulike.org http://www.citeulike.org/user/sona/article/2883797 <i>International Conference on Research in Electrotechnology and Applied Informatics (REI'05) (2005), pp. 109-119.</i><br /><br />A comparison between several feature ranking methods used on artificial and real dataset is presented. Six ranking methods based on entropy and statistical indices are considered. The Parzen window method for estimation of mutual information and other indices gives similar results as discretization based on the separability index, but results strongly dependent on the # smoothing parameter. The quality of the feature subsets with highest ranks is evaluated by using decision tree, Naive Bayes... Feature ranking methods based on information entropy with Parzen windows J Biesiada W Duch A Kachel K Maczka S Palucha International Conference on Research in Electrotechnology and Applied Informatics (REI'05) (2005), pp. 109-119. 2008-06-11T20:45:26-00:00 2005 International Conference on Research in Electrotechnology and Applied Informatics (REI'05) 109 119 artificial-data feature-ranking real-data survey http://www.citeulike.org/user/sona/article/2883773 <i>Knowledge Discovery in Databases: PKDD 2004 (2004), pp. 267-278.</i><br /><br />A crucial issue for Machine Learning and Data Mining is Feature Selection, selecting the relevant features in order to focus the learning search. A relaxed setting for Feature Selection is known as Feature Ranking, ranking the features with respect to their relevance. This paper proposes an ensemble approach for Feature Ranking, aggregating feature rankings extracted along independent runs of an evolutionary learning algorithm named ROGER. The convergence of ensemble feature ranking is studied in a theoretical perspective, and a statistical model is devised for the empirical validation, inspired from the complexity framework proposed in the Constraint Satisfaction domain. Comparative experiments demonstrate the robustness of the approach for learning (a limited kind of) non-linear concepts, specifically when the features significantly outnumber the examples. Ensemble Feature Ranking Kees Jong Jérémie Mary Antoine Cornuéjols Elena Marchiori Michèle Sebag Knowledge Discovery in Databases: PKDD 2004 (2004), pp. 267-278. 2008-06-11T20:31:46-00:00 2004 Knowledge Discovery in Databases: PKDD 2004 267 278 ensemble-methods feature-ranking filter-feature-selection http://www.citeulike.org/user/sona/article/2798974 <i>Vol. Tech. Rep. No. 1530 (December 2007)</i> Semi-Supervised Learning Literature Survey Xiaojin Zhu Vol. Tech. Rep. No. 1530 (December 2007) 2008-05-14T14:21:30-00:00 2007 Tech. Rep. No. 1530 semi-supervised survey http://www.citeulike.org/user/sona/article/2330827 <i>International Conference on Machine Learning (ICML) (2003), pp. 107-114.</i> On kernel methods for relational learning Chad Cumby Dan Roth International Conference on Machine Learning (ICML) (2003), pp. 107-114. 2008-02-04T20:40:56-00:00 2003 International Conference on Machine Learning (ICML) 107 114 composite-kernel kernel-methods relational-learning svm http://www.citeulike.org/user/sona/article/1202462 <i>International Conference on Computational Linguistics and the annual meeting of the ACL (2006), pp. 825-832.</i> A composite kernel to extract relations between entities with both flat and structured features Min Zhang Jie Zhang Jian Su Guodong Zhou doi:10.3115/1220175.1220279 International Conference on Computational Linguistics and the annual meeting of the ACL (2006), pp. 825-832. 2007-04-02T05:56:59-00:00 2006 International Conference on Computational Linguistics and the annual meeting of the ACL 825 832 Association for Computational Linguistics composite-kernel kernel-methods structured-input supervised svm http://www.citeulike.org/user/sona/article/315762 <i>Machine Learning, Vol. 24, No. 2. (1996), pp. 123-140.</i><br /><br />Bagging predictors is a method for generating multiple versions of a predictor and using these to get an aggregated predictor. The aggregation averages over the versions when predicting a numerical outcome and does a plurality vote when predicting a class. The multiple versions are formed by making bootstrap replicates of the learning set and using these as new learning sets. Tests on real and simulated data sets using classification and regression trees and subset selection in linear regression show that bagging can give substantial gains in accuracy. The vital element is the instability of the prediction method. If perturbing the learning set can cause significant changes in the predictor constructed, then bagging can improve accuracy. Bagging Predictors Leo Breiman doi:10.1023/A:1018054314350 Machine Learning, Vol. 24, No. 2. (1996), pp. 123-140. 2005-09-12T03:36:22-00:00 1996 Machine Learning 24 2 123 140 bagging survey http://www.citeulike.org/user/sona/article/2646542 <i>International Joint Conference on Neural Networks (IJCNN), Vol. 4 (2001), pp. 2478-2482 vol.4.</i><br /><br />This paper illustrates a new approach to sensitivity analysis for feature selection using multiple ensemble neural networks in a bootstrapping mode with bagging. This methodology is applied to in-silico drug design with QSAR (quantitative structural activity relationship), which is notoriously challenging for machine learning because typically there are on the order of 300-1000 dependent features, often for as few as 50-100 data points. For an HIV dataset with 160 wavelets descriptors, the number of relevant features was reduced to 35, and the resulting predictive neural network model gave better results than with the full feature set Bagging neural network sensitivity analysis for feature reduction for in-silico drug design MJ Embrechts F Arciniegas M Ozdemir CM Breneman K Bennett L Lockwood doi:10.1109/IJCNN.2001.938756 International Joint Conference on Neural Networks (IJCNN), Vol. 4 (2001), pp. 2478-2482 vol.4. 2008-04-09T17:00:11-00:00 2001 International Joint Conference on Neural Networks (IJCNN) 4 2478 2482 vol.4 dimensionality-reduction embedded-feature-selection feature-selection neural-networks sensitivity-analysis http://www.citeulike.org/user/sona/article/530829 <i>Journal of Machine Learning Research, Vol. 3 (2003), pp. 1229-1243.</i><br /><br />We describe a methodology for performing variable ranking and selection using support vector machines (SVMs). The method constructs a series of sparse linear SVMs to generate linear models that can generalize well, and uses a subset of nonzero weighted variables found by the linear models to produce a final nonlinear model. The method exploits the fact that a linear SVM (no kernels) with 1 -norm regularization inherently performs variable selection as a side-effect of minimizing capacity of the SVM model. The distribution of the linear model weights provides a mechanism for ranking and interpreting the effects of variables. Starplots are used to visualize the magnitude and variance of the weights for each variable. We illustrate the effectiveness of the methodology on synthetic data, benchmark problems, and challenging regression problems in drug design. This method can dramatically reduce the number of variables and outperforms SVMs trained using all attributes and using the attributes selected according to correlation coefficients. The visualization of the resulting models is useful for understanding the role of underlying variables. Dimensionality reduction via sparse support vector machines Jinbo Bi Kristin Bennett Mark Embrechts Curt Breneman Minghu Song Journal of Machine Learning Research, Vol. 3 (2003), pp. 1229-1243. 2006-03-04T11:22:02-00:00 2003 Journal of Machine Learning Research 1533-7928 3 1229 1243 MIT Press dimensionality-reduction embedded-feature-selection feature-selection kernel-methods regularization sparse-learning svm http://www.citeulike.org/user/sona/article/402254 <i>Artificial Intelligence, Vol. 97, No. 1. (December 1997), pp. 245-271.</i><br /><br /> In this survey, we review work in machine learning on methods for handling data sets containing large amounts of irrelevant information. We focus on two key issues: the problem of selecting relevant features, and the problem of selecting relevant examples. We describe the advances that have been made on these topics in both empirical and theoretical work in machine learning, and we present a general framework that we use to compare different methods. We close with some challenges for future work in this area. Selection of relevant features and examples in machine learning AL Blum P Langley doi:10.1016/S0004-3702(97)00063-5 Artificial Intelligence, Vol. 97, No. 1. (December 1997), pp. 245-271. 2005-11-20T23:23:42-00:00 1997 Artificial Intelligence 0004-3702 97 1 245 271 dimensionality-reduction embedded-feature-selection feature-selection filter-feature-selection survey weighting-feature-selection wrapper-feature-selection http://www.citeulike.org/user/sona/article/2646272 <i>Statistica Sinica, Vol. 16, No. 2. (2006), pp. 495-522.</i><br /><br />We present a Bayesian hierarchical model for indoor location estimation in wireless networks. We demonstrate that our model achieves accuracy that is similar to other published models and algorithms. By harnessing prior knowledge, our model drastically reduces the requirement for training data as compared with existing approaches. Location estimation in wireless networks: a Bayesian approach David Madigan Wen-Hua Ju P Krishnan AS Krishnakumar Ivan Zorych Statistica Sinica, Vol. 16, No. 2. (2006), pp. 495-522. 2008-04-09T15:35:27-00:00 2006 Statistica Sinica 16 2 495 522 bayesian-networks graphical-models localization monte-carlo sensor-networks http://www.citeulike.org/user/sona/article/2646107 <i>Neural Comput., Vol. 7, No. 2. (March 1995), pp. 219-269.</i> Regularization theory and neural networks architectures Federico Girosi Michael Jones Tomaso Poggio Neural Comput., Vol. 7, No. 2. (March 1995), pp. 219-269. 2008-04-09T15:18:23-00:00 1995 Neural Comput. 0899-7667 7 2 219 269 MIT Press neural-networks regularization http://www.citeulike.org/user/sona/article/1995614 <i>Statistica Sinica, Vol. 16 (2006), pp. 307-322.</i> Challenges in Statistical Machine Learning John Lafferty Larry Wasserman Statistica Sinica, Vol. 16 (2006), pp. 307-322. 2007-11-27T18:00:19-00:00 2006 Statistica Sinica 16 307 322 editorial high-dimensions relational-learning semi-supervised sparse-learning structured-prediction http://www.citeulike.org/user/sona/article/600433 <i>(2003)</i> Semi-Supervised Learning using Gaussian Fields and Harmonic Functions Xiaojin Zhu Zoubin Ghahramani John Lafferty (2003) 2006-04-25T16:02:08-00:00 2003 digits-recognition kernel-methods linked-input random-field real-dataset relational-learning semi-supervised http://www.citeulike.org/user/sona/article/2626199 <i>Journal of Machine Learning Research, Vol. 7 (July 2006), pp. 1159-1182.</i><br /><br />This paper introduces a learning method for two-layer feedforward neural networks based on sensitivity analysis, which uses a linear training algorithm for each of the two layers. First, random values are assigned to the outputs of the first layer; later, these initial values are updated based on sensitivity formulas, which use the weights in each of the layers; the process is repeated until convergence. Since these weights are learnt solving a linear system of equations, there is an important saving in computational time. The method also gives the local sensitivities of the least square errors with respect to input and output data, with no extra computational cost, because the necessary information becomes available without extra calculations. This method, called the Sensitivity-Based Linear Learning Method, can also be used to provide an initial set of weights, which significantly improves the behavior of other learning algorithms. The theoretical basis for the method is given and its performance is illustrated by its application to several examples in which it is compared with several learning algorithms and well known data sets. The results have shown a learning speed generally faster than other existing methods. In addition, it can be used as an initialization tool for other well known methods with significant improvements. A Very Fast Learning Method for Neural Networks Based on Sensitivity Analysis Enrique Castillo Bertha Guijarro-Berdiñas Oscar Fontenla-Romero Amparo Alonso-Betanzos Journal of Machine Learning Research, Vol. 7 (July 2006), pp. 1159-1182. 2008-04-03T14:18:05-00:00 2006 Journal of Machine Learning Research 7 1159 1182 fast-learning learning-algorithms neural-networks sensitivity-analysis supervised http://www.citeulike.org/user/sona/article/167555 <i>Journal of Machine Learning Research, Vol. 3 (2003), pp. 1157-1182.</i><br /><br />Variable and feature selection have become the focus of much research in areas of application for which datasets with tens or hundreds of thousands of variables are available. These areas include text processing of internet documents, gene expression array analysis, and combinatorial chemistry. The objective of variable selection is three-fold: improving the prediction performance of the pre- dictors, providing faster and more cost-effective predictors, and providing a better understanding of the underlying process that generated the data. The contributions of this special issue cover a wide range of aspects of such problems: providing a better definition of the objective function, feature construction, feature ranking, multivariate feature selection, efficient search methods, and feature validity assessment methods. An introduction to variable and feature selection Isabelle Guyon André Elisseeff Journal of Machine Learning Research, Vol. 3 (2003), pp. 1157-1182. 2005-04-22T17:16:32-00:00 2003 Journal of Machine Learning Research 1533-7928 3 1157 1182 MIT Press dimensionality-reduction feature-selection survey http://www.citeulike.org/user/sona/article/1530636 <i>J. Chem. Soc. Faraday Trans., Vol. 90, No. 1. (1994), pp. 97-102.</i> Use of Neural Network to determine the Boiling Point of Alkanes D Cherqaoui D Villemin J. Chem. Soc. Faraday Trans., Vol. 90, No. 1. (1994), pp. 97-102. 2007-08-02T13:10:04-00:00 1994 J. Chem. Soc. Faraday Trans. 90 1 97 102 chemistry neural-networks http://www.citeulike.org/user/sona/article/1530635 <i>Journal of Applied Intelligence (Kluwer Academic Publishers), Vol. 12 (2000), pp. 117-146.</i> Application of Cascade Correlation Networks for Structures to Chemistry AM Bianucci A Micheli A Sperduti A Starita Journal of Applied Intelligence (Kluwer Academic Publishers), Vol. 12 (2000), pp. 117-146. 2007-08-02T13:10:04-00:00 2000 Journal of Applied Intelligence (Kluwer Academic Publishers) 12 117 146 cascade-correlation chemistry structured-input http://www.citeulike.org/user/sona/article/1530634 <i>Neural Network, Vol. 14 (2001), pp. 1307-1321.</i> Bi-directional computing architectures for time series prediction H Wakuya J Zurada Neural Network, Vol. 14 (2001), pp. 1307-1321. 2007-08-02T13:10:04-00:00 2001 Neural Network 14 1307 1321 Elsevier cascade-correlation contextual-processing recurrent-neural-networks sequences structured-input http://www.citeulike.org/user/sona/article/1530633 <i>Neural Computation, Vol. 1 (1989), pp. 270-280.</i> A Learning Algorithm for Continually Running Fully Recurrent Neural Networks RJ Williams D Zipser Neural Computation, Vol. 1 (1989), pp. 270-280. 2007-08-02T13:10:04-00:00 1989 Neural Computation 1 270 280 MIT Press learning-algorithms recurrent-neural-networks http://www.citeulike.org/user/sona/article/1530632 <i>(2000)</i> Time Series Prediction by a Neural Network Model Based on the Bi-directional Computational Style H Wakuya J Zurada (2000) 2007-08-02T13:10:04-00:00 2000 contextual-processing neural-networks time-series http://www.citeulike.org/user/sona/article/1530631 <i>Vol. 3 (2000), pp. 3-8.</i> Bi-Causal Recurrent Cascade Correlation A Micheli D Sona A Sperduti Vol. 3 (2000), pp. 3-8. 2007-08-02T13:10:04-00:00 2000 3 3 8 cascade-correlation contextual-processing recurrent-neural-networks structured-input time-series http://www.citeulike.org/user/sona/article/1530630 <i>IEEE Transactions on Acoustics, Speech and Signal Processing, Vol. 37, No. 3. (1989), pp. 328-339.</i> Phoneme recognition using time--delay neural networks A Waibel T Hanazawa G Hinton K Shikano K Lang IEEE Transactions on Acoustics, Speech and Signal Processing, Vol. 37, No. 3. (1989), pp. 328-339. 2007-08-02T13:10:04-00:00 1989 IEEE Transactions on Acoustics, Speech and Signal Processing 37 3 328 339 bibtex-import http://www.citeulike.org/user/sona/article/1530629 <i>(1991), pp. 190-196.</i> The Recurrent Cascade-Correlation Architecture SE Fahlman (1991), pp. 190-196. 2007-08-02T13:10:04-00:00 1991 190 196 Morgan Kaufmann Publishers cascade-correlation recurrent-neural-networks http://www.citeulike.org/user/sona/article/1530628 <i>Bioinformatics, Vol. 15, No. 11. (1999), pp. 937-946.</i> Exploiting the past and the future in protein secondary structure prediction P Baldi S Brunak P Frasconi G Pollastri G Soda Bioinformatics, Vol. 15, No. 11. (1999), pp. 937-946. 2007-08-02T13:10:04-00:00 1999 Bioinformatics 15 11 937 946 bibtex-import http://www.citeulike.org/user/sona/article/1530627 <i>IEEE Transactions on Neural Networks, Vol. 8, No. 3. (1997), pp. 714-735.</i> Supervised Neural Networks for the Classification of Structures A Sperduti A Starita IEEE Transactions on Neural Networks, Vol. 8, No. 3. (1997), pp. 714-735. 2007-08-02T13:10:04-00:00 1997 IEEE Transactions on Neural Networks 8 3 714 735 neural-networks recursive-neural-networks structured-input http://www.citeulike.org/user/sona/article/1530626 <i>Journal of Molecular Biology, Vol. 202 (1988), pp. 865-884.</i> Predicting the secondary structure of globular proteins using neural network models N Qian TJ Sejnowski Journal of Molecular Biology, Vol. 202 (1988), pp. 865-884. 2007-08-02T13:10:04-00:00 1988 Journal of Molecular Biology 202 865 884 bibtex-import http://www.citeulike.org/user/sona/article/1530598 <i>Vol. 30 of Series in Information Sciences. (2001)</i> Self-Organizing Maps T Kohonen Vol. 30 of Series in Information Sciences. (2001) 2007-08-02T13:05:22-00:00 2001 30 of Series in Information Sciences. Springer centroid-based self-organizing-map unsupervised http://www.citeulike.org/user/sona/article/1530597 <i>Journal of Computing and Information Technology, Vol. 13, No. 4. (2005), pp. 279-285.</i> Simple classification into large topic ontology of Web documents M Grobelnik D Mladenic Journal of Computing and Information Technology, Vol. 13, No. 4. (2005), pp. 279-285. 2007-08-02T13:05:22-00:00 2005 Journal of Computing and Information Technology 13 4 279 285 bibtex-import http://www.citeulike.org/user/sona/article/1530596 <i>(1998)</i> Machine learning on non-homogeneus, distribuited text data D Mladenic (1998) 2007-08-02T13:05:22-00:00 1998 bibtex-import http://www.citeulike.org/user/sona/article/1530595 <i>Journal of Intelligent Data Analysis, Vol. 4, No. 5. (2000), pp. 411-420.</i> Intelligent document classification RA Calvo HA Ceccatto Journal of Intelligent Data Analysis, Vol. 4, No. 5. (2000), pp. 411-420. 2007-08-02T13:05:22-00:00 2000 Journal of Intelligent Data Analysis 4 5 411 420 bibtex-import http://www.citeulike.org/user/sona/article/1530594 <i>(1997), pp. 170-178.</i> Hierarchically Classifying Documents Using Very Few Words D Koller M Sahami (1997), pp. 170-178. 2007-08-02T13:05:22-00:00 1997 170 178 bibtex-import http://www.citeulike.org/user/sona/article/1530593 <i>Journal of ACM Computing Surveys, Vol. 34, No. 1. (2002), pp. 1-47.</i> Machine learning in automated text categorization F Sebastiani Journal of ACM Computing Surveys, Vol. 34, No. 1. (2002), pp. 1-47. 2007-08-02T13:05:22-00:00 2002 Journal of ACM Computing Surveys 34 1 1 47 bibtex-import http://www.citeulike.org/user/sona/article/1530592 <i>(1994), pp. 13-22.</i> Expert Network: Effective and Efficient Learning from Human Decisions in Text Categorization and Retrieval Y Yang (1994), pp. 13-22. 2007-08-02T13:05:22-00:00 1994 13 22 bibtex-import http://www.citeulike.org/user/sona/article/1530590 <i>(1999), pp. 52-58.</i> Text Classification by Bootstrapping with Keywords, EM and Shrinkage A Mccallum K Nigam (1999), pp. 52-58. 2007-08-02T13:05:22-00:00 1999 52 58 bibtex-import http://www.citeulike.org/user/sona/article/1530589 <i>(2003), pp. 57-72.</i> Web-pages Classification into a Hierarchy of Categories M Ceci D Malerba (2003), pp. 57-72. 2007-08-02T13:05:22-00:00 2003 57 72 bibtex-import http://www.citeulike.org/user/sona/article/1530588 <i>(2000), pp. 256-263.</i> Hierarchical classification of Web document S Dumais H Chen (2000), pp. 256-263. 2007-08-02T13:05:22-00:00 2000 256 263 bibtex-import http://www.citeulike.org/user/sona/article/1530587 <i>(2003)</i> Learning with taxonomies: Classifying documents and words T Hofmann L Caiand M Ciaramita (2003) 2007-08-02T13:05:22-00:00 2003 bibtex-import http://www.citeulike.org/user/sona/article/1530586 <i>(1999)</i> Building hierarchical classifiers using class proximity K Wang S Zhou S Liew (1999) 2007-08-02T13:05:22-00:00 1999 bibtex-import http://www.citeulike.org/user/sona/article/1530585 <i>(2006), pp. 177-184.</i> Hierarchical classification: combining Bayes with SVM Cesa Bianchi C Gentile L Zaniboni (2006), pp. 177-184. 2007-08-02T13:05:22-00:00 2006 177 184 bibtex-import http://www.citeulike.org/user/sona/article/1530584 <i>(2007)</i> Relational Cascade Correlation D Sona N Polettini (2007) 2007-08-02T13:05:22-00:00 2007 bibtex-import http://www.citeulike.org/user/sona/article/1530583 <i>(1999), pp. 52-58.</i> Text Classification by Bootstrapping with Keywords, EM and Shrinkage A Mccallum K Nigam (1999), pp. 52-58. 2007-08-02T13:05:22-00:00 1999 52 58 bibtex-import http://www.citeulike.org/user/sona/article/1530582 <i>(2003), pp. 295-302.</i> Bootstrapping for Hierarchical Document Classification G Adami P Avesani D Sona (2003), pp. 295-302. 2007-08-02T13:05:22-00:00 2003 295 302 bibtex-import http://www.citeulike.org/user/sona/article/1530581 <i>(2003), pp. 66-73.</i> Clustering Documents in a Web Directory G Adami P Avesani D Sona (2003), pp. 66-73. 2007-08-02T13:05:22-00:00 2003 66 73 bibtex-import http://www.citeulike.org/user/sona/article/1530580 <i>(2004), pp. 50-61.</i> Clustering with Propagation for Hierarchical Document Classification D Sona S Veeramachaneni P Avesani N Polettini (2004), pp. 50-61. 2007-08-02T13:05:22-00:00 2004 50 61 bibtex-import http://www.citeulike.org/user/sona/article/1530579 <i>(2005)</i> Helping Physicians to Organize Guidelines within Conceptual Hierarchies D Sona P Avesani R Moskovitch (2005) 2007-08-02T13:05:22-00:00 2005 bibtex-import http://www.citeulike.org/user/sona/article/1530578 <i>(2005), pp. 256-263.</i> Multi-Classification of Clinical Guidelines in Concept Hierarchies D Sona P Avesani R Moskovitch (2005), pp. 256-263. 2007-08-02T13:05:22-00:00 2005 256 263 bibtex-import http://www.citeulike.org/user/sona/article/1530577 <i>(2005), pp. 928-935.</i> Hierarchical Dirichlet Model for Document Classification S Veeramachaneni D Sona P Avesani (2005), pp. 928-935. 2007-08-02T13:05:22-00:00 2005 928 935 bibtex-import http://www.citeulike.org/user/sona/article/1530576 <i>(2006), pp. 691-696.</i> Regularization for Unsupervised Classification on Taxonomies D Sona S Veeramachaneni N Polettini P Avesani (2006), pp. 691-696. 2007-08-02T13:05:22-00:00 2006 691 696 bibtex-import http://www.citeulike.org/user/sona/article/1530575 <i>Journal of Data and Knowledge Engineering, Vol. 54, No. 1. (2005), pp. 301-325.</i> Clustering documents into a Web directory for bootstrapping a supervised classification G Adami P Avesani D Sona Journal of Data and Knowledge Engineering, Vol. 54, No. 1. (2005), pp. 301-325. 2007-08-02T13:05:22-00:00 2005 Journal of Data and Knowledge Engineering 54 1 301 325 bibtex-import http://www.citeulike.org/user/sona/article/1530574 <i>Scientific American, Vol. 236 (May 1977), pp. 119-127.</i> Stein's Paradox in Statistics B Efron C Morris Scientific American, Vol. 236 (May 1977), pp. 119-127. 2007-08-02T13:05:21-00:00 1977 Scientific American 236 119 127 bibtex-import