659 Hits in 5.7 sec

Gait verification using knee acceleration signals

L.W. Hang, C.Y. Hong, C.W. Yen, D.J. Chang, M.L. Nagurka
2011 Expert systems with applications  
A novel gait recognition method for biometric applications is proposed. The approach has the following distinct features.  ...  Second, an automatic procedure to extract gait features from acceleration signals is developed that employs a multipletemplate classification method.  ...  Without considering false acceptance errors, the above procedure achieves a zero false rejection error (and hence, 100% sensitivity) for the training samples.  ... 
doi:10.1016/j.eswa.2011.05.028 fatcat:26yuseg3nfbqrje2s5km6pneoy

A False Alarm Reduction Method for a Gas Sensor Based Electronic Nose

Mohammad Rahman, Chalie Charoenlarpnopparut, Prapun Suksompong, Pisanu Toochinda, Attaphongse Taparugssanagorn
2017 Sensors  
In addition, the MMM method was found to potentially reduce false classification errors.  ...  To reduce false classification and misclassification errors, and to improve correct rejection performance; algorithms with a hyperspheric boundary, such as a radial basis function neural network (RBFNN  ...  It is seen from the analysis as tabulated in Table 3 that the MMM method classified the test data of trained classes with only a 1.8519% misclassification error, i.e., false negative error.  ... 
doi:10.3390/s17092089 pmid:28895910 pmcid:PMC5620598 fatcat:43c5f46nzncydf7hdzpi6pbrpu

Ensemble Deep Learning for Cervix Image Selection toward Improving Reliability in Automated Cervical Precancer Screening

Peng Guo, Zhiyun Xue, Zac Mtema, Karen Yeates, Ophira Ginsburg, Maria Demarco, L. Rodney Long, Mark Schiffman, Sameer Antani
2020 Diagnostics  
The algorithm processes images of the uterine cervix taken with a digital camera and alerts the user if the woman is a candidate for further evaluation.  ...  In this work, we present a novel ensemble deep learning method to identify cervix images and non-cervix images in a smartphone-acquired cervical image dataset.  ...  RetinaNet Results: (a) ROC from one of the test folds, and (b) and false negative error examples.  ... 
doi:10.3390/diagnostics10070451 pmid:32635269 pmcid:PMC7400120 fatcat:lqvucs7o2vd7dnueclqil2l45m

Support Vector Data Description Model to Map Specific Land Cover with Optimal Parameters Determined from a Window-Based Validation Set

Jinshui Zhang, Zhoumiqi Yuan, Guanyuan Shuai, Yaozhong Pan, Xiufang Zhu
2017 Sensors  
The C is defined as the ratio of target objects to outlier objects in a training sample set and the kernel width s is to control the compactness of hypersphere.  ...  The support vector data description (SVDD) method, a boundary method developed by Tax and Duin, creates a hypersphere which is the decision boundary in a high-dimensional feature space such that it encloses  ...  We really appreciate the anonymous reviewers and editors' their valuable comments and suggestions for improving the quality of this paper.  ... 
doi:10.3390/s17050960 pmid:28445404 pmcid:PMC5461084 fatcat:urasuzssdjbjdenb6xxntklyt4

The Impact of Overfitting and Overgeneralization on the Classification Accuracy in Data Mining [chapter]

Huy Nguyen Anh Pham, Evangelos Triantaphyllou
2008 Soft Computing for Knowledge Discovery and Data Mining  
No single method outperforms all methods all the time. Furthermore, the performance of a classification method in terms of its false-positive and false-negative rates may be totally unpredictable.  ...  In this way, the previous three error rates can be controlled in a comprehensive manner.  ...  Case 4: Now we consider a case in which the application would penalize much more for the false-negative cases than for the other types of error.  ... 
doi:10.1007/978-0-387-69935-6_16 dblp:series/springer/PhamT08 fatcat:nahnhiqlwvhcbeoncqwob5ma7q

Genetic Algorithm for hardware Trojan detection with ring oscillator network (RON)

Nima Karimian, Fatemeh Tehranipoor, Md. Tauhidur Rahman, Shane Kelly, Domenic Forte
2015 2015 IEEE International Symposium on Technologies for Homeland Security (HST)  
The proposed method is an improvement over principal component analysis (PCA) in terms of accuracy and equal error rate by 30% and 97% respectively.  ...  Furthermore, we propose a novel feature selection approach based on the Genetic Algorithm (GA) and evaluate its performance compared to several popular alternatives.  ...  A ROC curve is illustrated in Figure 7 . This curve shows the trade-off between false acceptance rate (FAR) and true positive rate (TPR).  ... 
doi:10.1109/ths.2015.7225334 fatcat:rucbcchbmvftnguilxj3stj7hi

Sparse Kernel-Based Hyperspectral Anomaly Detection

P. Gurram, Heesung Kwon, T. Han
2012 IEEE Geoscience and Remote Sensing Letters  
In this letter, a novel ensemble-learning approach for anomaly detection is presented.  ...  In this method, the features of a given multivariate data set representing normalcy are first randomly subsampled into a large number of feature subspaces.  ...  of the hypersphere and the errors.  ... 
doi:10.1109/lgrs.2012.2187040 fatcat:5yg5om3pvrav3anr2fy5vqhrce

An Anomaly Detection Approach To Detect Unexpected Faults In Recordings From Test Drives

Andreas Theissler, Ian Dear
2013 Zenodo  
SVDDSUBSEQ is proposed as a novel approach, allowing to classify subsequences in multivariate time series data.  ...  The recordings are used for fault analysis. Since the resulting data volume is tremendous, manually analysing each recording in great detail is not feasible.  ...  The error on the normal class is minimised by adjusting R and a in a way that all instances of the training data set are contained in the hypersphere.  ... 
doi:10.5281/zenodo.1087118 fatcat:tqrhxmgnibgntfu2xvlr7moqgi

Discriminative Training for Multiple Observation Likelihood Ratio Based Voice Activity Detection

Tao Yu, John H L Hansen
2010 IEEE Signal Processing Letters  
In this study, the optimal combination weights from two discriminative training methods are studied to directly improve VAD performance, in terms of reduced misclassification errors and improved receiver  ...  However, the combination weights for the likelihood ratios (LRs) in each observation are rather empirical and heuristical.  ...  Here, the natural gradient is employed due to its optimality for the increasing direction on a hypersphere. IV. EVALUATIONS A.  ... 
doi:10.1109/lsp.2010.2066561 fatcat:pdicgeglgncy3kg7yega6agbmm

Spontaneous Emotional Facial Expression Detection

Zhihong Zeng, Yun Fu, Glenn I. Roisman, Zhen Wen, Yuxiao Hu, Thomas S. Huang
2006 Journal of Multimedia  
In this paper, we explore methods for detecting emotional facial expressions occurring in a realistic human conversation setting-the Adult Attachment Interview (AAI).  ...  Our preliminary experiments on AAI data suggest that one-class classification methods can reach a good balance between cost (labeling and computing) and recognition performance by avoiding non-emotional  ...  ACKNOWLEDGMENT We would like to thank Ms Debra Huang for discussion of her honor thesis.  ... 
doi:10.4304/jmm.1.5.1-8 fatcat:v3le6gw525f2fnekg5bgvtzr3q

Authenticity Detection of Black Rice by Near-Infrared Spectroscopy and Support Vector Data Description

Hui Chen, Chao Tan, Zan Lin
2018 International Journal of Analytical Chemistry  
On average, the optimized SVDD model achieves acceptable performance, i.e., a specificity of 100% and a sensitivity of 94.2% on the independent test set with tight boundary.  ...  There is increasing need for the development of fast, easy-to-use, and low-cost analytical methods for authenticity detection.  ...  , respectively; is the penalty factor which regulates the hyperspherical volume and error, i.e., the number of target objects rejected; is a slack variable for allowable error limitation.  ... 
doi:10.1155/2018/8032831 pmid:30105054 pmcid:PMC6076898 fatcat:w4nj5wjjdjh4lkdgb6n6sbtzzy

A meta-heuristic approach for improving the accuracy in some classification algorithms

Huy Nguyen Anh Pham, Evangelos Triantaphyllou
2011 Computers & Operations Research  
Furthermore, current algorithms ignore the fact that there may be different penalty costs for the false-positive, false-negative, and unclassifiable types.  ...  The CBA first defines the total misclassification cost (TC) as a weighted function of the three penalty costs and the corresponding error rates as mentioned above.  ...  Acknowledgements The authors are very appreciative for valuable comments made by the anonymous reviewers. These comments have helped the authors to improve the quality of this paper.  ... 
doi:10.1016/j.cor.2010.04.011 fatcat:eow4yrcfqnggplj4y4up5ph4ge

Effects of hyperellipsoidal decision surfaces on image segmentation in artificial color

H. John Caulfield
2010 Journal of Electronic Imaging (JEI)  
Of course, a hypersphere is just a degenerate hyperellipsoid; thus, exploring the effect of loosening that degeneracy seemed appropriate.  ...  Initially, we use two-foci hyperellipsoids with a hyperellipsoidal distance metric to classify pixels with dramatic improvement in performance.  ...  If those classifiers indicate a class, then it is accepted. If not, then we go to the second-stage classifiers, and so forth through all the predefined stages.  ... 
doi:10.1117/1.3377146 fatcat:qkbvdwpelbhwrpvagofxf3apgm

Intrusion Detection In Scada Systems Using One-Class Classification

Pierre Beauseroy, Paul Honeine, Patric Nader
2013 Zenodo  
the testbed: a) command injection attack where false control information is injected in the network traffic; b) response injection attack where false measurements are sent to the control system; and c)  ...  The reconstruction error defines a novelty measure. Fig. 1.  ... 
doi:10.5281/zenodo.43546 fatcat:47p6xiw3i5bptiqshy4of2km6y

Multi-Model Switching Based Fault Detection for the Suspension System of Maglev Train

Ping Wang, Zhiqiang Long, Ning Dang
2019 IEEE Access  
In order to satisfy the requirements of on-line monitoring of the suspension system under complex conditions, a fault detection method for maglev train suspension system based on multi-model switching  ...  The results demonstrate that the proposed method is superior to the other two methods in terms of health detection rate and false positive rate.  ...  The false positive rate F A2 of the second method and the false positive rate F A3 of the third method is 0.97% higher than the false alarm rate F A1 of the first method.  ... 
doi:10.1109/access.2018.2889733 fatcat:5o73hjtiujc3vci2vxtpbggqde
« Previous Showing results 1 — 15 out of 659 results