On output independence and complementariness in rank-based multiple classifier decision systems

Saranlı, Afşar; Demirekler, Mübeccel

doi:10.1016/s0031-3203(00)00175-8

Cited by 9 publications

(3 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Multiple classifier system trains multiple component classifiers and then fuses these decisions to approximate the target function from different views, which can significantly improve the generalization ability. Therefore, it has been a hot topic for the past 10 years and widely used in many areas, such as machineprinted or handwritten character recognition, image recognition, speaker recognition, face identification, biomedical signal processing and data mining [1,2] etc.…”

Section: Introductionmentioning

confidence: 99%

An improved multiple fuzzy NNC system based on mutual information and fuzzy integral

Wang

2010

Int. J. Mach. Learn. & Cyber.

View full text Add to dashboard Cite

Multiple nearest neighbor classifier system (MNNCS) is a popular method to relax the curse of dimensionality. In previous work, most of the MNNCSs are designed by random methods. Random methods may generate unstable component classifiers. In order to relax the randomness, large amount of component classifiers are needed. This paper first extends nearest neighbor classifier into fuzzy nearest neighbor classifier, and proposes a new multiple fuzzy nearest neighbor classifier system based on mutual information and fuzzy integral, called MIFI-MFNNCS. MIFI-MFNNCS adopts target perturbation. Target perturbation decomposes the original classification problem into several sub-problems, where one sub-problem represents one class data. Each sub-problem is described by the relevant data and features. Then it is classified by one component classifier. Therefore, the number of component classifiers can be fixed and reduced. For one component classifier, data may be selected according to its class. And feature is needed to be selected by mutual information. Mutual information can reduce the uncertainty of each component classifier. Feature selection by mutual information in MIFI-MFNNCS may be less affected by the interaction among different classes. The diversity decisions from sub-problem classifiers are combined by fuzzy integral to get the final decision. Here we propose a new method to compute density value according to mutual information, which is a simple method. To demonstrate the performance of the proposed MIFI-MFNNCS, we perform experimental comparisons using five UCI datasets. The results of component classifiers in MIFI-MFNNCS for Ionosphere are shown and analyzed. MIFI-MFNNCS is compared with (1) NNC (2) NNC after feature selection by mutual information (MI-FS-NNC). In multiple fuzzy nearest neighbor classifier system (MFNNCS), mutual information is compared with attribute bagging. And three combination methods are compared, including fuzzy integral, majority voting rule and average. The experimental results show that the accuracy of MIFI-MFNNCS is better than other methods. And mutual information is superior to attribute bagging. Fuzzy integral shows a better performance than majority voting rule and average.

show abstract

Section: Introductionmentioning

confidence: 99%

An improved multiple fuzzy NNC system based on mutual information and fuzzy integral

Wang

2010

Int. J. Mach. Learn. & Cyber.

View full text Add to dashboard Cite

show abstract

“…They used a unified (general) statistical approach based on the partitioned observation space (POS) theory. Specific partitioning of the classifier observation space lead to the highest rank, Borda count or logistic regression rankbased combination methods [14] [13]. Nandakumar et al [10] proposed a Bayesian approach for rank fusion.…”

Section: Introductionmentioning

confidence: 99%

Quality based rank-level fusion in multibiometric systems

Abaza

Ross

2009

2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems

View full text Add to dashboard Cite

Multibiometric systems fuse evidences from multiple biometric sources typically resulting in better recognition accuracy. These systems can consolidate information at various levels. For systems operating in the identification mode, rank level fusion presents a viable option. In this paper, several simple but powerful modifications are suggested to enhance the performance of rank-level fusion schemes in the presence of weak classifiers or low quality input images. These modifications do not require a training phase, therefore making them suitable in a wide range of applications. Experiments conducted on a multimodal database consisting of a few hundred users indicate that the suggested modifications to the highest rank and Borda count methods significantly enhance the rank-1 accuracy. Experiments also reveal that including image quality in the fusion scheme enhances the Borda count rank-1 accuracy by ∼ 40%.

show abstract

“…Information theory and in particular mutual information had been applied to several machine learning problems, such as modeling of self organized systems and feature maps [25,9], feature transformation and selection [7,40], image processing [8,41], independent component analysis [14], evaluation of the relations between output independence and complementariness in multiple classifier decision systems [36].…”

Section: Introductionmentioning

confidence: 99%

An experimental analysis of the dependence among codeword bit errors in ECOC learning machines

Masulli

Valentini

2004

Neurocomputing

View full text Add to dashboard Cite

One of the main factors affecting the effectiveness of Error Correcting Output Coding (ECOC) methods for classification is the dependence among the errors of the computed codeword bits. We present an extensive experimental work for evaluating the dependence among output errors of the decomposition unit in ECOC learning machines. In particular, we apply measures based on mutual information to compare the dependence of ECOC Multi-Layer Perceptron (ECOC MLP), made up by a single multi-input multi-output MLP, and ECOC ensembles made up by a set of independent and parallel dichotomizers (ECOC PND). Moreover, the experimentation analyzes the relationship between the architecture, the dependence among output errors and the performances of ECOC learning machines. The results show that the dependence among computed codeword bits is significantly smaller for ECOC PND, pointing out that ensembles of independent parallel dichotomizers are better suited for implementing ECOC classification methods. The experimental results suggest new architectures of ECOC learning machines to improve the independence among output errors and the diversity between base learners.

show abstract

On output independence and complementariness in rank-based multiple classifier decision systems

Cited by 9 publications

References 28 publications

An improved multiple fuzzy NNC system based on mutual information and fuzzy integral

An improved multiple fuzzy NNC system based on mutual information and fuzzy integral

Quality based rank-level fusion in multibiometric systems

An experimental analysis of the dependence among codeword bit errors in ECOC learning machines

Contact Info

Product

Resources

About