Classifiers Combination Techniques: A Comprehensive Review
作者： Mohamed Mohandes , Mohamed Deriche , Salihu O. Aliyu
DOI: 10.1109/ACCESS.2018.2813079
关键词: Computer science 、 Curse of dimensionality 、 Statistical classification 、 Artificial intelligence 、 Classifier (UML) 、 Sensor fusion 、 Machine learning 、 Medical diagnosis 、 Feature extraction 、 Thresholding
摘要: In critical applications, such as medical diagnosis, security related systems, and so on, the cost or risk of action taking based on incorrect classification can be very high. Hence, combining expert opinions before decision substantially increase reliability systems. Such pattern recognition systems base their final evidence collected from different classifiers. data type, feature classifier type. Common problems in recognition, curse dimensionality, small sample size, among others, have also prompted researchers into seeking new approaches for evidences. This paper presents a criteria-based framework multi-classifiers combination techniques areas applications. The criteria discussed here include levels combination, types thresholding, adaptiveness ensemble-based approaches. strengths weaknesses each these categories are details. Following this analysis, we provide our perspective outlook area research open problems. lack well-formulated theoretical analyzing performance is shown to fertile ground further research. addition summarizing existing work, updates complements latest developments
sci-hub.se 参考文章(74)
Mohammed Abdul Azeem Siddiqui, Fusion of ECG/EEG for improved automatic seizure detection using Dempster Shafer theory of evidence ,(2011)
Gaurav Jain, A. Ginwala, Y.A. Asiandogan, An approach to text classification using dimensionality reduction and combination of classifiers information reuse and integration. pp. 564- 569 ,(2004) , 10.1109/IRI.2004.1431521
Sergey Tulyakov, Stefan Jaeger, Venu Govindaraju, David Doermann, Review of Classifier Combination Methods Machine Learning in Document Analysis and Recognition. pp. 361- 386 ,(2008) , 10.1007/978-3-540-76280-5_14
Josef Kittler, A Framework for Classifier Fusion: Is It Still Needed? Lecture Notes in Computer Science. pp. 45- 56 ,(2000) , 10.1007/3-540-44522-6_5
Umit Ekmekci, Zehra Cataltepe, Classifier combination with kernelized eigenclassifiers international conference on information fusion. pp. 743- 749 ,(2013)
Imran Naseem, Combining Classifiers Using the Dempster Shafer Theory of Evidence ,(2010)
K. Sirlantzis, M.C. Fairhurst, Optimisation of multiple classifier systems using genetic algorithms international conference on image processing. ,vol. 1, pp. 1094- 1097 ,(2001) , 10.1109/ICIP.2001.959240
V. Di Lecce, G. Dimauro, A. Guerriero, S. Impedovo, G. Pirlo, A. Salzo, Knowledge-based methods for classifier combination: an experimental investigation international conference on image analysis and processing. pp. 562- 565 ,(1999) , 10.1109/ICIAP.1999.797655
A. Rahman, M. Fairhurst, Decision combination of multiple classifiers for pattern classification: hybridisation of majority voting and divide and conquer techniques workshop on applications of computer vision. pp. 58- 63 ,(2000) , 10.1109/WACV.2000.895403
J.J. de Oliveira, M.N. Kapp, C.O.A. Freitas, J.M. de Carvalho, R. Sabourin, Handwritten recognition with multiple classifiers for restricted lexicon brazilian symposium on computer graphics and image processing. pp. 82- 89 ,(2004) , 10.1109/SIBGRA.2004.1352947