Applying knowledge engineering and representation methods to improve support vector machine and multivariate probabilistic neural network CAD performance

2005 ◽  
Author(s):  
Walker H. Land, Jr. ◽  
Frances Anderson ◽  
Tom Smith ◽  
Stephen Fahlbusch ◽  
Robert Choma ◽  
...  
Keyword(s):  
Neural Network ◽  
Support Vector Machine ◽  
Knowledge Engineering ◽  
Probabilistic Neural Network ◽  
Support Vector

AUTOMATED RECOGNITION OF SINGLE & HYBRID POWER QUALITY DISTURBANCES USING WAVELET TRANSFORM BASED SUPPORT VECTOR MACHINE

2016 ◽  
Vol 79 (1) ◽  
Author(s):  
Suhail Khokhar ◽  
A. A. Mohd Zin ◽  
M. A. Bhayo ◽  
A. S. Mokhtar
Keyword(s):  
Neural Network ◽  
Support Vector Machine ◽  
Wavelet Transform ◽  
Power Quality ◽  
Probabilistic Neural Network ◽  
Recognition Rate ◽  
Classification Performance ◽  
Support Vector ◽  
Discrete Wavelet ◽  
Feed Forward Neural Network

The monitoring of power quality (PQ) disturbances in a systematic and automated way is an important issue to prevent detrimental effects on power system. The development of new methods for the automatic recognition of single and hybrid PQ disturbances is at present a major concern. This paper presents a combined approach of wavelet transform based support vector machine (WT-SVM) for the automatic classification of single and hybrid PQ disturbances. The proposed approach is applied by using synthetic models of various single and hybrid PQ signals. The suitable features of the PQ waveforms were first extracted by using discrete wavelet transform. Then SVM classifies the type of PQ disturbances based on these features. The classification performance of the proposed algorithm is also compared with wavelet based radial basis function neural network, probabilistic neural network and feed-forward neural network. The experimental results show that the recognition rate of the proposed WT-SVM based classification system is more accurate and much better than the other classifiers. 

scholarly journals Multi-Stage Feature Selection (MSFS) Algorithm for UWB-Based Early Breast Cancer Size Prediction

2020 ◽  
Author(s):  
V. Vijayasarveswari ◽  
A.M. Andrew ◽  
M. Jusoh ◽  
T. Sabapathy ◽  
R.A.A. Raof ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Neural Network ◽  
Support Vector Machine ◽  
Feature Selection ◽  
Probabilistic Neural Network ◽  
The Other ◽  
Ultra Wideband ◽  
Support Vector ◽  
Data Normalization ◽  
Multi Stage

AbstractBreast cancer is the most common cancer among women and it is one of the main causes of death for women worldwide. To attain an optimum medical treatment for breast cancer, an early breast cancer detection is crucial. This paper proposes a multistage feature selection method that extracts statistically significant features for breast cancer size detection using proposed data normalization techniques. Ultra-wideband (UWB) signals, controlled using microcontroller are transmitted via an antenna from one end of the breast phantom and are received on the other end. These ultra-wideband analogue signals are represented in both time and frequency domain. The preprocessed digital data is passed to the proposed multi-stage feature selection algorithm. This algorithm has four selection stages. It comprises of data normalization methods, feature extraction, data dimensional reduction and feature fusion. The output data is fused together to form the proposed datasets, namely, 8-HybridFeature, 9-HybridFeature and 10-HybridFeature datasets. The classification performance of these datasets is tested using the Support Vector Machine, Probabilistic Neural Network and Naïve Bayes classifiers for breast cancer size classification. The research findings indicate that the 8-HybridFeature dataset performs better in comparison to the other two datasets. For the 8-HybridFeature dataset, the Naïve Bayes classifier (91.98%) outperformed the Support Vector Machine (90.44%) and Probabilistic Neural Network (80.05%) classifiers in terms of classification accuracy. The finalized method is tested and visualized in the MATLAB based 2D and 3D environment.

scholarly journals Fault Diagnosis Method for Engine Control System Based on Probabilistic Neural Network and Support Vector Machine

2019 ◽  
Vol 9 (19) ◽  
pp. 4122 ◽  
Author(s):  
Bo Wang ◽  
Hongwei Ke ◽  
Xiaodong Ma ◽  
Bing Yu
Keyword(s):  
Neural Network ◽  
Support Vector Machine ◽  
Control System ◽  
Fault Diagnosis ◽  
Probabilistic Neural Network ◽  
Support Vector ◽  
Engine Control ◽  
Piston Engine ◽  
Engine Control System ◽  
Diagnosis Method

Due to the poor working conditions of an engine, its control system is prone to failure. If these faults cannot be treated in time, it will cause great loss of life and property. In order to improve the safety and reliability of an aero-engine, fault diagnosis, and optimization method of engine control system based on probabilistic neural network (PNN) and support vector machine (SVM) is proposed. Firstly, using the German 3 W piston engine as a control object, the fault diagnosis scheme is designed and introduced briefly. Then, the fault injection is performed to produce faults, and the data sample for engine fault diagnosis is established. Finally, the important parameters of PNN and SVM are optimized by particle swarm optimization (PSO), and the results are analyzed and compared. It shows that the engine fault diagnosis method based on PNN and SVM can effectively diagnose the common faults. Under the optimization of PSO, the accuracy of PNN and SVM results are significantly improved, the classification accuracy of PNN is up to 96.4%, and the accuracy of SVM is up to 98.8%, which improves the application of them in fault diagnosis technology of aero-piston engine control system.

scholarly journals Evaluation of Performance of Decision Tree, Support Vector Machine and Probabilistic Neural Network Classifiers in a Mobile Based Diabetes Retinopathy Detection System

2019 ◽  
pp. 1-9
Author(s):  
O. D. Fenwa ◽  
O. O. Alo ◽  
I. O. Omotoso
Keyword(s):  
Neural Network ◽  
Support Vector Machine ◽  
Diabetic Retinopathy ◽  
Decision Tree ◽  
Medical Condition ◽  
Detection System ◽  
Probabilistic Neural Network ◽  
Rejection Rate ◽  
Support Vector ◽  
Diabetes Retinopathy

Diabetic Retinopathy (DR) is a medical condition where the retina is damaged because fluid leaks from blood vessels into the retina. Ophthalmologists recognize diabetic retinopathy based on features, such as blood vessel area, exudes, hemorrhages, microaneurysms and texture. Aim: The focus of this paper is to evaluate the performance of Decision Tree (DT), Support Vector Machine (SVM) and Probabilistic Neural Network (PNN) Classifiers in Diabetes Retinopathy Detection. Results: Corresponding results showed SVM has the best classification strength by achieving Recognition Accuracy (RA) of 98.50%, while PNN and DT achieved RA of 97.60% and 89.20% respectively. In terms of False Acceptance Rate (FAR) and False Rejection Rate (FRR), SVM has the least values of 7.21, 8.10 while DT and PNN showed 11.10, 9.30 and 13.21, 10.10 respectively. However, in this paper a Mobile based Diabetes Retinopathy Detection System was developed to make the system available for the masses for early detection of the disease.

Quantity Prediction of Mixed Freshwater Fish Based on Passive Underwater Acoustic Signals

2020 ◽  
Vol 63 (6) ◽  
pp. 1805-1811
Author(s):  
Qunzi Tu ◽  
Yongwen Yang ◽  
Hanying Huang ◽  
Lu Li ◽  
Shanbai Xiong ◽  
...  
Keyword(s):  
Neural Network ◽  
Support Vector Machine ◽  
Linear Regression ◽  
Prediction Model ◽  
Freshwater Fish ◽  
Probabilistic Neural Network ◽  
Recognition Rate ◽  
Acoustic Signals ◽  
Support Vector ◽  
Underwater Acoustic

HighlightsThe use of passive underwater acoustic technology to estimate the species and quantity of freshwater fish provides a theoretical basis for effectively estimating the quantity of freshwater aquaculture.Mixed proportion recognition models for breams and crucians were built using probabilistic neural network (PNN) and support vector machine (SVM), and the influences of different super-parameters on the recognition rate were analyzed. The results showed that the classification model established with SVM after equiripple filtering was best.Mixed quantity prediction models for breams and crucians were constructed using multiple linear regression, and the effects of different filtering methods on the model performance were analyzed. The results showed that the best quantity prediction model was constructed with Butterworth filtering.Abstract. Acoustic signals of breams and crucians were collected at seven mixed proportions and 15 mixed quantitative gradients. After normalization and different filtering processes, the characteristics of the acoustic signals were extracted. Mixed proportion recognition models for breams and crucians were established using probabilistic neural network (PNN) and support vector machine (SVM). The results showed that the model established using SVM after equiripple filtering was best, and the recognition rate was 0.9583. A mixed quantity prediction model for breams and crucians was established by multiple linear regression based on ordinary least squares. The results showed that the model was best after Butterworth filtering, the adjusted decision coefficient of the model was 0.9514, and the relative analysis error was 4.7571. Keywords: Freshwater fish, Passive underwater acoustic signals, Pattern recognition, Regression analysis.

Sign in / Sign up

Export Citation Format

Share Document