A Comparative Study of Improved F-Score with Support Vector Machine and RBF Network for Breast Cancer Classification

Jaganathan, P.; Rajkumar, N.; Kuppuchamy, R.

doi:10.7763/ijmlc.2012.v2.227

Cited by 14 publications

(8 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In particular, we employed the Fisher score [60] selection strategy, which calculates the ratio of interclass separation and intraclass variance for each feature and thus expresses the discriminative power of each feature [61,64]. The Fisher score is one of the most widely used supervised feature selection methods and has been combined with SVM classifications in many recent studies [62,[65][66][67][68]. In detail, the Fisher score of the rth feature is computed by the formula…”

Section: Methodsmentioning

confidence: 99%

Detailed Land Cover Mapping from Multitemporal Landsat-8 Data of Different Cloud Cover

et al. 2018

View full text Add to dashboard Cite

Detailed, accurate and frequent land cover mapping is a prerequisite for several important geospatial applications and the fulfilment of current sustainable development goals. This paper introduces a methodology for the classification of annual high-resolution satellite data into several detailed land cover classes. In particular, a nomenclature with 27 different classes was introduced based on CORINE Land Cover (CLC) Level-3 categories and further analysing various crop types. Without employing cloud masks and/or interpolation procedures, we formed experimental datasets of Landsat-8 (L8) images with gradually increased cloud cover in order to assess the influence of cloud presence on the reference data and the resulting classification accuracy. The performance of shallow kernel-based and deep patch-based machine learning classification frameworks was evaluated. Quantitatively, the resulting overall accuracy rates differed within a range of less than 3%; however, maps produced based on Support Vector Machines (SVM) were more accurate across class boundaries and the respective framework was less computationally expensive compared to the applied patch-based deep Convolutional Neural Network (CNN). Further experimental results and analysis indicated that employing all multitemporal images with up to 30% cloud cover delivered relatively higher overall accuracy rates as well as the highest per-class accuracy rates. Moreover, by selecting 70% of the top-ranked features after applying a feature selection strategy, slightly higher accuracy rates were achieved. A detailed discussion of the quantitative and qualitative evaluation outcomes further elaborates on the performance of all considered classes and highlights different aspects of their spectral behaviour and separability.

show abstract

Section: Methodsmentioning

confidence: 99%

Detailed Land Cover Mapping from Multitemporal Landsat-8 Data of Different Cloud Cover

et al. 2018

View full text Add to dashboard Cite

show abstract

“…Based on Fig. 4, all nonzero singular values are used to determine the weights between hidden layer and output layer (14). Table I, the highest accuracy for training and testing data is achieved at 81.1% and 91.7%, respectively for ten clusters.…”

Section: Resultsmentioning

confidence: 99%

Diagnosis of Brain Cancer Using Radial Basis Function Neural Network with Singular Value Decomposition Method

Abadi¹,

Wustqa²,

Nurhayadi³

2019

IJMLC

View full text Add to dashboard Cite

Brain cancer is one of the most dangerous cancers that can attack anyone, so early detection needs to be done so that brain cancer can be treated quickly. The purpose of this study is to develop a new procedure of modeling radial basis function neural network (RBFNN) using singular value decomposition (SVD) method and to apply the procedure to diagnose brain cancer. This study uses 114 brain Magnetic Resonance Images (MRI). The RBFNN model is constructed by using steps as follows; image preprocessing, image extracting using Gray Level Co-Occurrence Matrix (GLCM), determining of parameters of radial basis function and determining of weights of neural network. The input variables are 14 features of image extractions and the output variable is a classification of brain cancer. Before learning process, the input data is normalized. The modeling is done by using K-means clustering method where the activation function in the hidden layer is Gaussian function and by determining the optimum weights of the model using SVD method. The best RBFNN model is 14 input neurons, 10 hidden layer neurons, and 1 output neuron. The results show that the sensitivity, specificity, and accuracy of RBFNN diagnoses with backpropagation equal to those of RBFNN with SVD. However, the RBFNN-SVD delivers an advantage in the running speed of the program.Index Terms-Radial basis function neural network, diagnosis brain cancer, singular value decomposition.

show abstract

“…where -is the coefficient of the predictor variable and slope can be interpreted as the change of , from unit change in [10]. The LR model can be expressed as follows:…”

Section: Logistic Regression (Lr)mentioning

confidence: 99%

“…Varies application motivated by the success of the classification techniques, especially in the medical domain [8]- [10] utilized widely. Therefore, an objective of these designed built to compute the classifiers evaluation, in the result, explore the best models for supporting their decision.…”

Section: Introductionmentioning

confidence: 99%

Prediction of 6 Months Smoking Cessation Program among Women in Korea

Davagdorj¹,

Yu²,

Kim³

et al. 2019

IJMLC

View full text Add to dashboard Cite

Cigarette smoking is the leading cause of preventable death in a general population and it seems a significant topic in health research. The primary aim of this study determines the significant risk factors and investigates the prediction of 6 months smoking cessation program among women in Korea. In this regard, we examined real-world dataset about a smoking cessation program among the only women from Chungbuk Tobacco Control Center of Chungbuk National University College of Medicine in South Korea which collected from 2015 to 2017. Accordingly, we carried out to compare four machine learning techniques: Logistic regression (LR), Support Vector Machine (SVM), Random Forest (RF) and Naï ve Bayes (NB) in order to predict response for successful or unsuccessful smoking quitters. Totally we analyzed 60 set of features that may affect the association between smoking cessation such as socio-demographic characteristics, smoking status for the age of starting, duration and others by employing a filter-based feature selection method. Respectively, we identified significant 8 factors which associated with smoking cessation. The experimental results demonstrate that NB performs better than other classifiers. Moreover, the performance of prediction models as measured by Accuracy, Precision, Recall, F-measure and ROC area. This finding has gone some way towards enhancing our better understanding of the significant factors contributing to smoking cessation program implementation and accompanying to concern public health.

show abstract

A Comparative Study of Improved F-Score with Support Vector Machine and RBF Network for Breast Cancer Classification

Cited by 14 publications

References 11 publications

Detailed Land Cover Mapping from Multitemporal Landsat-8 Data of Different Cloud Cover

Detailed Land Cover Mapping from Multitemporal Landsat-8 Data of Different Cloud Cover

Diagnosis of Brain Cancer Using Radial Basis Function Neural Network with Singular Value Decomposition Method

Prediction of 6 Months Smoking Cessation Program among Women in Korea

Contact Info

Product

Resources

About