Filters








287,005 Hits in 4.4 sec

Bagging and the Random Subspace Method for Redundant Feature Spaces [chapter]

Marina Skurichina, Robert P. W. Duin
2001 Lecture Notes in Computer Science  
In this paper, on the example of the pseudo Fisher linear classifier, we study the effect of the redundancy in the data feature set on the performance of the random subspace method and bagging. b  ...  The performance of a single weak classifier can be improved by using combining techniques such as bagging, boosting and the random subspace method.  ...  Acknowledgment This work is supported by the Foundation for Applied Sciences (STW) and the Dutch Organization for Scientific Research (NWO).  ... 
doi:10.1007/3-540-48219-9_1 fatcat:nbj72qlwgbf4bjojwnq2azxl6y

A linear discriminant analysis framework based on random subspace for face recognition

Xiaoxun Zhang, Yunde Jia
2007 Pattern Recognition  
The two sets of discriminant analysis features from dual principal subspaces are first combined at the feature level, and then all the random subspaces are further integrated at the decision level.  ...  Under the most suitable situation of the principal subspace, the optimal reduced dimension of the face sample is discovered to construct a random subspace where all the discriminative information in the  ...  on the feature level and the decision level resp ectively, our framewo rk exploits more discriminant power in the face space with the discriminating information fusion at the two levels.  ... 
doi:10.1016/j.patcog.2006.12.002 fatcat:rpfbezfzxrfbfa4mnirwaelg3a

Exploring the Best Classification from Average Feature Combination

Jian Hou, Wei-Xue Liu, Hamid Reza Karimi
2014 Abstract and Applied Analysis  
In previous work we have found that it is better to use only a sample of the most powerful features in average combination than using all.  ...  Furthermore, the kNN framework is helpful in understanding the underlying mechanism of feature combination and motivating novel feature combination algorithms.  ...  In Section 2 we attribute this observation to the variance of the discriminative power of individual features.  ... 
doi:10.1155/2014/602763 fatcat:xjxswmkva5c5pj3cvbqnvvbinu

Directed Random Subspace Method for Face Recognition

Mehrtash T. Harandi, Majid Nili Ahmadabadi, Babak N. Araabi, Abbas Bigdeli, Brian C. Lovell
2010 2010 20th International Conference on Pattern Recognition  
In this paper we present a learning scheme to overcome some of the drawbacks of random feature selection in the random subspace method.  ...  The proposed learning method derives a feature discrimination map based on a measure of accuracy and uses it in a probabilistic recall mode to construct an ensemble of subspaces.  ...  Element (i, j) of this table represents the expected discrimination power of selecting feature j right after selecting feature i.  ... 
doi:10.1109/icpr.2010.659 dblp:conf/icpr/HarandiAABL10 fatcat:maow7vmo2zavlgbnnlxec3jriu

Model Selection Framework for Graph-based data [article]

Rajmonda S. Caceres, Leah Weiner, Matthew C. Schmidt, Benjamin A. Miller, William M. Campbell
2016 arXiv   pre-print
We fully characterize the discriminative power of our approach as we sweep through the parameter space of two generative models, the Erdos-Renyi and the stochastic block model.  ...  We show that our approach gets very close to known theoretical bounds and we provide insight on which topological features play a critical discriminating role.  ...  However, using only ten features seems to greatly affect the discrimination power of the classifier.  ... 
arXiv:1609.04859v1 fatcat:7jw4uncnnzhmjpzo3wqg2rwkqe

Random forest to differentiate dementia with Lewy bodies from Alzheimer's disease

Meenakshi Dauwan, Jessica J. van der Zande, Edwin van Dellen, Iris E.C. Sommer, Philip Scheltens, Afina W. Lemstra, Cornelis J. Stam
2016 Alzheimer's and Dementia: Diagnosis, Assessment and Disease Monitoring  
Beta power was identified as the single-most important discriminating variable. qEEG increased the accuracy of the other multimodal diagnostic data with almost 10%.  ...  Results: For discrimination between DLB and AD, the diagnostic accuracy of the classifier was 87%.  ...  The authors declared no conflict of interest. 1  ... 
doi:10.1016/j.dadm.2016.07.003 pmid:27722196 pmcid:PMC5050257 fatcat:g7fgzgwsjndjfp3euage7r7nha

Efficient Discriminative K-SVD for Facial Expression Recognition [chapter]

Weifeng Liu, Caifeng Song, Yanjiang Wang
2013 Lecture Notes in Electrical Engineering  
To tackle this problem, we employ random projection on Gabor features and then put the reduced features into D-KSVD schema to obtain sparse representation and dictionary.  ...  Discriminative K-SVD (D-KSVD) is one of conventional dictionary learning methods, which can effectively unify dictionary learning and classifier.  ...  of K-SVD combining the representation power of K-SVD and discriminate ability of linear classifier.  ... 
doi:10.1007/978-3-642-38466-0_2 fatcat:icmh3n5dnjghjnatp5h6bdy6vm

Automated identification of cell-type–specific genes and alternative promoters [article]

Mickaël Mendez, Jayson Harshbarger, Michael M. Hoffman
2021 bioRxiv   pre-print
For each cell type pair, CLA runs multiple instances of regularized random forest and reports the transcriptional features consistently selected.  ...  CLA not only discriminates individual cell types but can also discriminate lineages of cell types related in the developmental hierarchy.  ...  This work was supported by the Natural Sciences and Engineering Research Council of Canada (RGPIN-2015-03948 to M.M.H.).  ... 
doi:10.1101/2021.12.01.470587 fatcat:vjy42wvvszhmhly77zwdkmag5a

Machine Learning Seismic Wave Discrimination: Application to Earthquake Early Warning

Zefeng Li, Men-Andrin Meier, Egill Hauksson, Zhongwen Zhan, Jennifer Andrews
2018 Geophysical Research Letters  
We apply the GAN critic as an automatic feature extractor and train a Random Forest classifier with about 700,000 earthquake and noise waveforms.  ...  We show that the discriminator can recognize 99.2% of the earthquake P waves and 98.4% of the noise signals.  ...  The Japanese waveform data can be downloaded Ti 4 GB, Intel Core i5-7300HQ 2.50GHz).  ... 
doi:10.1029/2018gl077870 fatcat:ydwi42itmfdejhodkgzpr3zib4

Nearest neighbors in random subspaces [chapter]

Tin Kam Ho
1998 Lecture Notes in Computer Science  
Recent studies have shown that the random subspace method can be used to create multiple independent tree-classifiers that can be combined to improve accuracy.  ...  We examine the effects of several parameters of the method by experiments using data from a digit recognition problem.  ...  The slower increases in accuracy are due to weaker discriminative power of component classifiers using a smaller number of features.  ... 
doi:10.1007/bfb0033288 fatcat:iljrpryjqvcafnx3h25gtrlba4

Learning feature extractors for AMD classification in OCT using convolutional neural networks

Dafydd Ravenscroft, Jingjing Deng, Xianghua Xie, Louise Terry, Tom H. Margrain, Rachel V. North, Ashley Wood
2017 2017 25th European Signal Processing Conference (EUSIPCO)  
kernels, and then generalizes the discriminative power by forming a histogram based descriptor.  ...  The experimental results show that the proposed method extracts more discriminative features than the features learnt through CNN only.  ...  Feature Generalization CNNs are usually powerful discriminators but we introduce histograms for feature generalization.  ... 
doi:10.23919/eusipco.2017.8081167 dblp:conf/eusipco/RavenscroftDXTM17 fatcat:uxwiuncy6bgijjdcozhdf7xhvq

Building Emerging Pattern (EP) Random forest for recognition

Liang Wang, Yizhou Wang, Debin Zhao
2010 2010 IEEE International Conference on Image Processing  
We employ an efficient data mining algorithm, the Emerging Pattern (EP) Mining, to search such discriminative patterns and weight them according to their discriminative powers.  ...  By treating the outputs of the tree classifiers on each data as a digital itemset, we want to find discriminative patterns (either containing the output of a single tree classifier or a set of tree classifiers  ...  The decision rules are weighted according to their discriminative power and the final decision of the Random Forest classifier is made by aggregating the votes of these decision rules, and these votes  ... 
doi:10.1109/icip.2010.5653902 dblp:conf/icip/WangWZ10 fatcat:a2y44kn6drd7jcwv6oan4lvvey

Towards mobile authentication using dynamic signature verification: Useful features and performance evaluation

Marcos Martinez-Diaz, Julian Fierrez, Javier Galbally, Javier Ortega-Garcia
2008 Pattern Recognition (ICPR), Proceedings of the International Conference on  
In this paper we study the effects of the mobile acquisition conditions and we analyze the considerations that must be taken in the new handheld scenario.  ...  The proliferation of handheld devices such as PDAs and smartphones represents a new scenario for automatic signature verification.  ...  An analysis of the discriminant power of different types of features (temporal, geometric, etc.) is performed using the Fisher Discriminant Ratio (FDR) and feature selection algorithms.  ... 
doi:10.1109/icpr.2008.4761849 dblp:conf/icpr/Martinez-DiazFGO08 fatcat:l37nmginjvbibedstqts5ib42y

Ordinal Random Forests for Object Detection [chapter]

Samuel Schulter, Peter M. Roth, Horst Bischof
2013 Lecture Notes in Computer Science  
In this way, we can also exploit more than two feature dimensions, resulting in increased discriminative power.  ...  Recent works showed that such statistics have more discriminative power than just observing single feature dimensions.  ...  ., image classification) and show the discriminative power of this non-linear transformation.  ... 
doi:10.1007/978-3-642-40602-7_29 fatcat:5oddsefcarfhjemtksyxppak34

Binary plankton image classification using random subspace

Feng Zhao, Xiaoou Tang, Feng Lin, S. Samson, A. Remsen
2005 IEEE International Conference on Image Processing 2005  
Using random sampling, we construct a set of stable classifiers to take full advantage of nearly all the discriminative information in the feature space of plankton images.  ...  In this paper, we implement a random subspace based algorithm to classify the plankton images detected in real time by the Shadowed Image Particle Profiling and Evaluation Recorder.  ...  ACKNOWLEDGMENT The work described in this paper was fully supported by grants from the Research Grants Council of the Hong Kong Special Administrative Region, and by the Office of Naval Research under  ... 
doi:10.1109/icip.2005.1529761 dblp:conf/icip/ZhaoTLSR05 fatcat:icwirzt5gfedlkuf6meim5b5s4
« Previous Showing results 1 — 15 out of 287,005 results