35,402 Hits in 3.1 sec

Soft Methodology for Cost-and-error Sensitive Classification [article]

Te-Kang Jan, Da-Wei Wang, Chi-Hung Lin, Hsuan-Tien Lin
2017 arXiv   pre-print
The methodology, called soft cost-sensitive classification, is established from a multicriteria optimization problem of the cost and the error rate, and can be viewed as regularizing cost-sensitive classification  ...  Many real-world data mining applications need varying cost for different types of classification errors and thus call for cost-sensitive classification algorithms.  ...  The proposed methodology, soft cost-sensitive classification, takes both the cost and the error (or the weighted error) into account by a multicriteria optimization problem.  ... 
arXiv:1710.09515v1 fatcat:zwk3exijszbf7ng77elkj37obe

A simple methodology for soft cost-sensitive classification

Te-Kang Jan, Da-Wei Wang, Chi-Hung Lin, Hsuan-Tien Lin
2012 Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '12  
The methodology, called soft cost-sensitive classification, is established from a multicriteria optimization problem of the cost and the error rate, and can be viewed as regularizing cost-sensitive classification  ...  Many real-world data mining applications need varying cost for different types of classification errors and thus call for cost-sensitive classification algorithms.  ...  The authors thank Yao-Nan Chen, Ku-Chun Chou, Chih-Han Yu and the anonymous reviewers for valuable comments.  ... 
doi:10.1145/2339530.2339555 dblp:conf/kdd/JanWLL12 fatcat:vzgmkndv7fbrpmxviflkouv6by

A New Cost Function for Binary Classification Problems Based on the Distributions of the Soft Output for Each Class

Marcelino Lazaro, Jose M. Leiva-Murillo, Antonio Artes-Rodriguez, Anibal R. Figueiras-Vidal
2007 Neural Networks (IJCNN), International Joint Conference on  
This cost function aims at reducing the probability of classification error by reducing the overlap between distributions of the soft output for each class.  ...  This paper proposes a new cost function for supervised training of neural networks in binary classification applications.  ...  TIC2003-02602, and project 'MONIN', id. TEC2006-13514-C02-01), and Comunidad de Madrid (project 'PRO-MULTIDIS-CM', id. S0505/TIC/0223).  ... 
doi:10.1109/ijcnn.2007.4371174 dblp:conf/ijcnn/LazaroLAF07 fatcat:qzrqtolnznfprd2huwnoqtcbky

Subspace electrode selection methodology for the reduction of the effect of uncertain conductivity values in the EEG dipole localization: a simulation study using a patient-specific head model

G Crevecoeur, V Montes Restrepo, S Staelens
2012 Physics in Medicine and Biology  
This paper presents a novel numerical methodology for the increase of accuracy of the EEG dipole source localization problem.  ...  We redefine the traditional cost function associated with the EEG inverse problem and introduce a selection procedure of EEG potentials.  ...  Acknowledgments The authors wish to thank Hans Hallez and Bertrand Yitembe, and gratefully acknowledge the financial support of GOA07/GOA/006 and IUAP P6/21.  ... 
doi:10.1088/0031-9155/57/7/1963 pmid:22421525 fatcat:7n4ogehj7vfobhjs42suh72dda

Analyzing the instructions vulnerability of dense convolutional network on GPUS

Khalid Adam, Izzeldin I. Mohd, Younis Ibrahim
2021 International Journal of Power Electronics and Drive Systems (IJPEDS)  
Our results show that the most significant vulnerable instructions against soft errors PR, STORE, FADD, FFMA, SETP and LD can be reduced from 4.42% to 0.14% of injected faults, after we applied our mitigation  ...  However, the GPUs have been reportedly impacted by soft errors, which are extremely serious issues in the healthcare applications.  ...  CNN model DenseNet201 was used GPUs with logic units more sensitive to soft errors was used.  ... 
doi:10.11591/ijece.v11i5.pp4481-4488 fatcat:sjvfjemysrf2nmfodixdwwvz4i

Analyzing the Impact of Soft Errors in Deep Neural Networks on GPUsfrom Instruction Level

2020 WSEAS transactions on systems and control  
In this work, we propose a comprehensive methodology to analyze the reliability of object detection and classification algorithms run on GPUs from the lowest (instruction) evaluation level.  ...  Moreover, we show that YOLO is more sensitive to the changes caused by soft errors than ResNet. Also, ResNet depends on the input image in its reliability, while YOLO tends to be independent.  ...  through SASSIFI fault injection to char acterize the error resilience behaviors of im age classification and object detection systems against soft errors. • Error criticality measurement to classify instruc  ... 
doi:10.37394/23203.2020.15.70 fatcat:cchrrxeozfduxpoccts7pxz2ie

SBT soft fault diagnosis in analog electronic circuits: a sensitivity-based approach by randomized algorithms

C. Alippi, M. Catelani, A. Fort, M. Mugnaini
2002 IEEE Transactions on Instrumentation and Measurement  
The method, based on a harmonic analysis, allows for selecting the most suitable test input stimuli and nodes by means of a global sensitivity approach efficiently carried out by randomized algorithms.  ...  Diagnosis, obtained by comparing signatures measured at the test nodes with those contained in a fault dictionary, allows for sub-systems testing and fault isolation within the circuit.  ...  The classification error improvement in fault detection and isolation with respect to the optimized procedure suggested in [14] is around 2%, which represents a relevant increment due to the cost of  ... 
doi:10.1109/tim.2002.806004 fatcat:jwip5w4w2nghvjg5lvpwjxateq


D. Aju, R. Rajkumar
2016 Jurnal Teknologi  
Magnetic Resonance Imaging (MRI) T1-weighted and T2-weighted images provides suitable variation of contrast between the different soft tissues of the brain which is suitable for detecting the brain tumor  ...  The proposed method uses the Regularized Logistic Regression (RLR) for the efficient cataloguing of brain tumor in which it achieves an effective accuracy rate of 96%, specificity rate of 86% and sensitivity  ...  It is interestingly noted that the proposed methodology, WATS+ RLR was able to achieve 96% of accuracy, 86 % of specificity and 97% of sensitivity when executed.  ... 
doi:10.11113/jt.v78.5930 fatcat:kj6ncxxagfhvjokxzekon6v5nu

Entropy-Based Modeling for Estimating Soft Errors Impact on Binarized Neural Network Inference [article]

Navid Khoshavi, Saman Sargolzaei, Arman Roohi, Connor Broyles, Yu Bi
2020 arXiv   pre-print
These models can be used for evaluating the error-resiliency magnitude of NN topology before adopting them in the safety-critical applications.  ...  In this paper, we present the relatively-accurate statistical models to delineate the impact of both undertaken SEU and MBU across layers and per each layer of the selected NN.  ...  To optimize the architecture design for ML/DL accelerators, a broad spectrum of methodologies have been proposed at both software and hardware levels.  ... 
arXiv:2004.05089v2 fatcat:j6j3gcjarrffzgecjqndfmg76q

Rejection Threshold Optimization using 3D ROC Curves: Novel Findings on Biomedical Datasets

Asli Uyar
2021 International Journal of Intelligent Systems and Applications in Engineering  
Although optimization of error-reject trade-off has been widely investigated, it is shown that error rate itself is not an appropriate performance measure, when misclassification costs are unequal or class  ...  Considering classification with reject option, we need to represent the tradeoff between TP, FP and rejection rates.  ...  He obtained Bayesian optimal errorreject characteristics by using cost sensitive classification and changing the decision threshold.  ... 
doi:10.18201/ijisae.2021167933 fatcat:zgk5n3eoavfyvoxnmogxnkrsuy

An Adaptive Resource Allocating Neuro-Fuzzy Inference System with Sensitivity Analysis Resource Control [chapter]

Minas Pertselakis, Natali Raouzaiou, Andreas Stafylopatis
2009 IFIP Advances in Information and Communication Technology  
In this framework, we present a novel methodology of dynamic resource control and optimization for neurofuzzy inference systems.  ...  Adaptability in non-stationary contexts is a very important property and a constant desire for modern intelligent systems and is usually associated with dynamic system behaviors.  ...  Introduction The guiding principle of soft computing is to exploit the tolerance for imprecision by devising methods of computation that lead to an acceptable solution at low cost [1] .  ... 
doi:10.1007/978-1-4419-0221-4_59 fatcat:xj37puf7o5fpzokd422mrkrpcy

Population Based Meta Heuristics Algorithm for Performance Improvement of Feed Forward Neural Network

Dr. Samuel Manoharan, Prof. Sathish
2020 Journal of Soft Computing Paradigm  
campaigns, research process and the scientific endeavors.  ...  Multitudes of methods were devised to improve the accuracy in the classification to devour an enhanced performance in terms of faster convergence speed.  ...  , Sensitivity and Error Rate Where the accuracy is measured using the[ 𝐴 𝐵 ⁄ ], sensitivity is estimated applying[ 𝑡 𝑝 𝐶 ⁄ ], and applying[ 𝑡 𝑛 𝐷 ⁄ ], for specificity whereas 1-[ 𝐴 𝐵 ⁄ ] for  ... 
doi:10.36548/jscp.2020.1.004 fatcat:cgkxov3uyzeh7og5ljkknom4jy

Support Vector Machines Classification on Class Imbalanced Data: A Case Study with Real Medical Data

Krystallenia Drosou, Stelios Georgiou, Christos Koukouvinos, Stella Stylianou
2021 Journal of Data Science  
We compare the standard (or classic) SVM (C SVM) with resampling methods and a cost sensitive method, called Two Cost SVM (TC SVM), which constitute widely accepted strategies for imbalanced datasets and  ...  In our comparative study we use a cost sensitive learning technique proposed by Veropoulos et al. (1999) called "TC SVM" due to the fact that it uses two costs for the two different classes.  ...  Acknowledgements The authors would like to thank the Editor and the Referee for their valuable comments and suggestions that resulted in improving the quality of presentation of this manuscript.  ... 
doi:10.6339/jds.201410_12(4).0009 fatcat:4s3xqbh5h5g7xhdfhj6qrf445i

Coronary Artery Disease Diagnosis Using Optimized Adaptive Ensemble Machine Learning Algorithm

Burak Kolukisa, Department of Computer Engineering, Abdullah Gül University, Kayseri, Turkey, Levent Yavuz, Ahmet Soran, Bakir-Gungor Burcu, Dilsad Tuncer, Ahmet Onen, V. Cagri Gungor
2020 International Journal of Bioscience Biochemistry and Bioinformatics  
In this study, we have proposed a novel Self Optimized and Adaptive Ensemble Machine Learning Algorithm for the diagnosis of CAD.  ...  Cardiovascular diseases (CVD) involving the heart and blood vessels are reported as the leading causes of mortality worldwide.  ...  Author Contributions All authors had conducted the research and approved the final version.  ... 
doi:10.17706/ijbbb.2020.10.1.58-65 fatcat:bfprjzqrmravlhi7lcwns4nhnu

Intelligent Decision Support System for Breast Cancer [chapter]

R. R. Janghel, Anupam Shukla, Ritu Tiwari, Rahul Kala
2010 Lecture Notes in Computer Science  
In this paper we develop an integrated expert system for diagnosis, prognosis and prediction for breast cancer using soft computing techniques.  ...  Breast cancer database used for this purpose is from the University of Wisconsin (UCI) Machine Learning Repository.  ...  Deshmukh, Director ABV-IIITM, Gwalior, India for encouraging and providing facilities to carry out this research work. This work is sponsored by ABV-IIITM Gwalior.  ... 
doi:10.1007/978-3-642-13498-2_46 fatcat:i6atgfy3vbhlfbnu3nryssb5i4
« Previous Showing results 1 — 15 out of 35,402 results