Filters








1,693 Hits in 9.6 sec

Error rate control for classification rules in multiclass mixture models [article]

Tristan Mary-Huard
2021 arXiv   pre-print
In the context of finite mixture models one considers the problem of classifying as many observations as possible in the classes of interest while controlling the classification error rate in these same  ...  is defined as minimizing type II error rate while controlling type I error rate at some nominal level.  ...  In Section 2, notions corresponding to type I and type II error rates are defined in the multiclass classification framework along with optimality for a classification rule.  ... 
arXiv:2109.14235v1 fatcat:anuzfnuf4fhu7g5ryierqtrwua

Interpretable Multiclass Models for Corporate Credit Rating Capable of Expressing Doubt

Lennart Obermann, Stephan Waack
2016 Frontiers in Applied Mathematics and Statistics  
Furthermore, Thresholder marks many potential misclassifications in advance with a doubt label without increasing the classification error.  ...  Furthermore, credit rating often is a multiclass problem with more than two rating classes.  ...  In this study, we examine interpretable multiclass models for a three-class credit rating.  ... 
doi:10.3389/fams.2016.00016 fatcat:lkckkvxorbbwbguq52caqkohyq

Performance-Based Classifier Combination in Atlas-Based Image Segmentation Using Expectation-Maximization Parameter Estimation

T. Rohlfing, D.B. Russakoff, C.R. Maurer
2004 IEEE Transactions on Medical Imaging  
The two methods are multiclass extensions of an expectation-maximization (EM) algorithm for ground truth estimation of binary classification based on decisions of multiple experts (Warfield et al., 2004  ...  The conventional method for combining individual classifiers weights each classifier equally (vote or sum rule fusion).  ...  ACKNOWLEDGMENT All computations were performed on an SGI Origin 3800 supercomputer in the Stanford University Bio-X Core Facility for Biomedical Computation.  ... 
doi:10.1109/tmi.2004.830803 pmid:15338732 fatcat:e6ye62rmjfbthpratd6it3kd5i

Class Proportion Estimation with Application to Multiclass Anomaly Rejection [article]

Tyler Sanderson, Clayton Scott
2014 arXiv   pre-print
This property allows us to address the second domain adaptation problem, namely, multiclass anomaly rejection.  ...  We establish consistent learning strategies for both of these domain adaptation problems, which to our knowledge are the first of their kind.  ...  Scott was supported in part by NSF Grants 0953135, 1047871, and 1217880. Note we truncated the sizes of some multiclass datasets in order to process them in a timely manner.  ... 
arXiv:1306.5056v3 fatcat:hou5hm7sljhgvbkjbbftiw4mb4

Multinomial probit Bayesian additive regression trees

Bereket P. Kindo, Hao Wang, Edsel A. Peña
2016 Stat  
Through two simulation studies and four real data examples, we show that MPBART exhibits very good predictive performance in comparison to other discrete choice and multiclass classification methods.  ...  George for his 2012 Palmetto Lecture at the University of South Carolina, which partly motivated this research.  ...  The authors also thank Professor James Lynch and Professor Edsel Peña's research group (A.K.M Rahman, Lillian Wanda, Piaomu Liu) for their comments and discussions.  ... 
doi:10.1002/sta4.110 pmid:27330743 pmcid:PMC4909838 fatcat:57glvv7e3rhyffbmybjdecuk3a

Learning Multiclass Rules with Class-Selective Rejection and Performance Constraints [chapter]

Nisrine Jrad, Pierre Beauseroy, Edith Grall-Maes
2010 Pattern Recognition Recent Advances  
the error rate and no constraint is considered.  ...  In a bayesian framework minimizing the error rate loss, estimating a complete data density for each of the N classes might be too demanding when only the data boundary is required.  ...  /books/pattern-recognition-recentadvances/learning-multiclass-rules-with-class-selective-rejection-and-performance-constraints  ... 
doi:10.5772/9360 fatcat:qxym5tht2bhuxelxpffnwogf5i

Selective sampling algorithms for cost-sensitive multiclass prediction

Alekh Agarwal
2013 International Conference on Machine Learning  
In this paper, we study the problem of active learning for cost-sensitive multiclass classification.  ...  For these algorithms, we analyze the regret and label complexity when the labels are generated according to a generalized linear model.  ...  and Francesco Orabona for helpful discussions.  ... 
dblp:conf/icml/Agarwal13 fatcat:pt2s7kl5jnfyja7fh53zaoaf3q

MPBART - Multinomial Probit Bayesian Additive Regression Trees [article]

Bereket P. Kindo, Hao Wang, Edsel A. Peña
2016 arXiv   pre-print
Through two simulation studies and four real data examples, we show that MPBART exhibits very good predictive performance in comparison to other discrete choice and multiclass classification methods.  ...  George for his 2012 Palmetto Lecture at the University of South Carolina, which partly motivated this research.  ...  Classification error rates (with standard errors in parentheses) are reported. . Multiclass Classification Example Datasets.  ... 
arXiv:1309.7821v2 fatcat:qzjwz4ckufhddlvtuczlh2oyq4

Digit recognition in noisy environments via a sequential GMM/SVM system

Fine, Saon, Gopinath
2002 IEEE International Conference on Acoustics Speech and Signal Processing  
The gain accrues from combining the descriptive strength of GMM models with the discriminative power of SVM classifiers.  ...  This idea, first exploited in the context of speaker recognition [1, 2] , is applied to speech recognition -specifically to a digit recognition task in a noisy environment -with significant gains in performance  ...  INTRODUCTION State-of-the-art speech recognizers use Hidden Markov Models with continuous features where each state is modeled by a Gaussian mixture.  ... 
doi:10.1109/icassp.2002.1005672 fatcat:oyke2htd6ba5jjfodaj6ixhbqy

Digit recognition in noisy environments via a sequential GMM/SVM system

Shai Fine, George Saon, Ramesh A. Gopinath
2002 IEEE International Conference on Acoustics Speech and Signal Processing  
The gain accrues from combining the descriptive strength of GMM models with the discriminative power of SVM classifiers.  ...  This idea, first exploited in the context of speaker recognition [1, 2] , is applied to speech recognition -specifically to a digit recognition task in a noisy environment -with significant gains in performance  ...  INTRODUCTION State-of-the-art speech recognizers use Hidden Markov Models with continuous features where each state is modeled by a Gaussian mixture.  ... 
doi:10.1109/icassp.2002.5743651 dblp:conf/icassp/FineSG02 fatcat:rz7rna733je37jjduevc6o2g44

Blind source separation and feature extraction in concurrent control charts pattern recognition: Novel analyses and a comparison of different methods

Guilherme Dean Pelegrina, Leonardo Tomazeli Duarte, Christian Jutten
2016 Computers & industrial engineering  
Control charts are among the main tools in statistical process control (SPC) and have been extensively used for monitoring industrial processes.  ...  the selection of efficient separation methods is fundamental to achieve high classification rates.  ...  Acknowledgements The Authors thank FAPESP and CNPq for funding their research projects.  ... 
doi:10.1016/j.cie.2015.12.017 fatcat:htnxiztq4bfapjpevnjv6p6orm

Fast neural network algorithm for solving classification tasks: Batch error back-propagation algorithm

Noor Albarakati, Vojislav Kecman
2013 2013 Proceedings of IEEE Southeastcon  
= + ′� � (3.13) Since LMS is applied in EBP algorithm, which is the most popular algorithm for multiclass classification problems, it is better to specify the error signal term δ for the OL in the formula  ...  to solve multiclass classification tasks.  ... 
doi:10.1109/secon.2013.6567409 fatcat:2eyr4vxxmne77iftklvbd4bloe

Fault diagnosis via structural support vector machines

Yi Peng, Qixiang Ye, Jianbin Jiao, Xiaogang Chen, Lijun Wu
2012 2012 IEEE International Conference on Mechatronics and Automation  
We define error penalty function and select a proper kernel to make structural SVMs be appropriate for non-linear problem.  ...  Discriminative methods are becoming more and more popular on fault diagnosis systems, while they need additional strategies or multiple models to cope with the multiple classification problems.  ...  We train a model for predicting the multiclass test data, rather than multiple models.  ... 
doi:10.1109/icma.2012.6284371 fatcat:s3zmer4v6ndtrdaxfpy54wqszu

Kernel Logistic Regression and the Import Vector Machine

Ji Zhu, Trevor Hastie
2005 Journal of Computational And Graphical Statistics  
The support vector machine (SVM) is known for its good performance in two-class classification, but its extension to multiclass classification is still an ongoing research issue.  ...  We show that the IVM not only performs as well as the SVM in two-class classification, but also can naturally be generalized to the multiclass case.  ...  We are also grateful for the three reviewers and one associate editor for their comments that helped improve the article. Ji Zhu was partially supported by the Stanford Graduate Fellowship.  ... 
doi:10.1198/106186005x25619 fatcat:wg5sd3d6zbcqnbyd2ymaf4g64e

Feasibility of Machine Learning Algorithms for Predicting the Deformation of Anodic Titanium Films by Modulating Anodization Processes

Sung-Hee Kim, Chanyoung Jeong
2021 Materials  
In contrast, decision tree and three ensemble methods had a relatively higher performance for multiclass classification, with an accuracy rate greater than 0.79.  ...  We applied eight machine learning techniques to predict classification for binary and multiclass classification.  ...  Multiclass Classification Table 4 . 4 Comparison of evaluation metrics for classification algorithm on multiclass classification.  ... 
doi:10.3390/ma14051089 pmid:33652708 pmcid:PMC7956670 fatcat:xgbrzi5i7zdcvkbey2hic7j43u
« Previous Showing results 1 — 15 out of 1,693 results