scholarly journals Applying WPD and SVD to Classification of EM Wave Induced by Partial Discharge in Power Transformer

2013 ◽  
Vol 64 (4) ◽  
pp. 222-229 ◽  
Author(s):  
Xu Zhao ◽  
Yong-Hong Cheng ◽  
Yong-Peng Meng ◽  
Michael G. Danikas
Keyword(s):  
Power Transformer ◽  
Wavelet Packet ◽  
Partial Discharge ◽  
Training Data ◽  
Decomposition Level ◽  
Frequency Information ◽  
Testing Data ◽  
The Right ◽  
Wave Induced

Partial discharge (PD) current is an impulse signal at nanosecond level, which can generate electromagnetic (EM) wave containing broadband frequency information. The frequency band of EM signal is from MHz up to GHz. Due to different PD patterns, impulse currents with different shapes induce different EM waves containing different frequency information. Therefore, using the features extracted from frequency domain of EM signals, the classification of PD patterns can be effectively got. It is good to use wavelet or wavelet packet decomposition to select features. However, if the decomposition level is too shallow to find enough effective features, it cannot group the EM signals to the right pattern. On the contrary, although it is easier to find features to distinguish the PD pattern if the decomposition level is deep, there will be a lot of redundancy variables and it is hard to select features among so many variables. In this paper, a method is presented, which selected features in the whole decomposition tree instead of selecting among the leaf node of the tree, because more potential features can be found in the whole tree. With the present method, it is possible not only to get enough features, but also to eliminate the redundancy variables effectively. In order to validate the method, large EM signals from four PD patterns in a power transformer are acquired as the training data and testing data for feature selection and classification, and three common classification methods are introduced to classify the PD patterns using the features selected by the method. Most of the classification results are satisfactory indicating that the proposed method is effective.

Classification of Breast Thermograms Using Statistical Moments and Entropy Features with Probabilistic Neural Networks

2020 ◽  
pp. 1328-1340
Author(s):  
Natarajan Sriraam ◽  
Leema Murali ◽  
Amoolya Girish ◽  
Manjunath Sirur ◽  
Sushmitha Srinivas ◽  
...  
Keyword(s):  
Probabilistic Neural Network ◽  
Wavelet Packet ◽  
Histogram Equalization ◽  
Developed Countries ◽  
Left Breast ◽  
Statistical Moments ◽  
Thermal Images ◽  
Life Threatening ◽  
The Right

Breast cancer is considered as one of the life-threatening disease among woman population in developing as well as developed countries. This specific study reports on classification of breast thermograms using probabilistic neural network (PNN) with four statistical moments features mean, standard deviation, skewness and kurtosis and two entropy features, Shannon entropy and Wavelet packet entropy. The CLAHE histogram equalization algorithm with uniform and Rayleigh distributions were considered for contrast enhancement of breast thermal images. The asymmetry detection was performed by applying bilateral ratio. A total of 95 test images (normal = 53, abnormal = 42) was considered. Simulation study shows that CLAHE -RD with wavelet entropy features confirms the existence of symmetry on the right and left breast thermal images. An overall classification accuracy of 92.5% was achieved using the proposed multifeatures with PNN classifier. The proposed technique thus confirms the suitability as a screening tool for asymmetry detection as well as classification of breast thermograms.

Classification of Breast Thermograms Using Statistical Moments and Entropy Features with Probabilistic Neural Networks

2020 ◽  
pp. 1175-1187
Author(s):  
Natarajan Sriraam ◽  
Leema Murali ◽  
Amoolya Girish ◽  
Manjunath Sirur ◽  
Sushmitha Srinivas ◽  
...  
Keyword(s):  
Probabilistic Neural Network ◽  
Wavelet Packet ◽  
Histogram Equalization ◽  
Developed Countries ◽  
Left Breast ◽  
Statistical Moments ◽  
Thermal Images ◽  
Life Threatening ◽  
The Right

Breast cancer is considered as one of the life-threatening disease among woman population in developing as well as developed countries. This specific study reports on classification of breast thermograms using probabilistic neural network (PNN) with four statistical moments features mean, standard deviation, skewness and kurtosis and two entropy features, Shannon entropy and Wavelet packet entropy. The CLAHE histogram equalization algorithm with uniform and Rayleigh distributions were considered for contrast enhancement of breast thermal images. The asymmetry detection was performed by applying bilateral ratio. A total of 95 test images (normal = 53, abnormal = 42) was considered. Simulation study shows that CLAHE -RD with wavelet entropy features confirms the existence of symmetry on the right and left breast thermal images. An overall classification accuracy of 92.5% was achieved using the proposed multifeatures with PNN classifier. The proposed technique thus confirms the suitability as a screening tool for asymmetry detection as well as classification of breast thermograms.

Classification of Breast Thermograms Using Statistical Moments and Entropy Features with Probabilistic Neural Networks

2017 ◽  
Vol 6 (2) ◽  
pp. 18-32
Author(s):  
Natarajan Sriraam ◽  
Leema Murali ◽  
Amoolya Girish ◽  
Manjunath Sirur ◽  
Sushmitha Srinivas ◽  
...  
Keyword(s):  
Probabilistic Neural Network ◽  
Wavelet Packet ◽  
Histogram Equalization ◽  
Developed Countries ◽  
Left Breast ◽  
Statistical Moments ◽  
Thermal Images ◽  
Life Threatening ◽  
The Right

Breast cancer is considered as one of the life-threatening disease among woman population in developing as well as developed countries. This specific study reports on classification of breast thermograms using probabilistic neural network (PNN) with four statistical moments features mean, standard deviation, skewness and kurtosis and two entropy features, Shannon entropy and Wavelet packet entropy. The CLAHE histogram equalization algorithm with uniform and Rayleigh distributions were considered for contrast enhancement of breast thermal images. The asymmetry detection was performed by applying bilateral ratio. A total of 95 test images (normal = 53, abnormal = 42) was considered. Simulation study shows that CLAHE -RD with wavelet entropy features confirms the existence of symmetry on the right and left breast thermal images. An overall classification accuracy of 92.5% was achieved using the proposed multifeatures with PNN classifier. The proposed technique thus confirms the suitability as a screening tool for asymmetry detection as well as classification of breast thermograms.

scholarly journals KLASIFIKASI CITRA DIGITAL BUMBU DAN REMPAH DENGAN ALGORITMA CONVOLUTIONAL NEURAL NETWORK (CNN)

2020 ◽  
Vol 9 (3) ◽  
pp. 273-282
Author(s):  
Isna Wulandari ◽  
Hasbi Yasin ◽  
Tatik Widiharih
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Activation Function ◽  
Training Data ◽  
Young Generation ◽  
Testing Data ◽  
Filter Size ◽  
Hidden Layer ◽  
Herbs And Spices

The recognition of herbs and spices among young generation is still low. Based on research in SMK 9 Bandung, showed that there are 47% of students that did not recognize herbs and spices. The method that can be used to overcome this problem is automatic digital sorting of herbs and spices using Convolutional Neural Network (CNN) algorithm. In this study, there are 300 images of herbs and spices that will be classified into 3 categories. It’s ginseng, ginger and galangal. Data in each category is divided into two, training data and testing data with a ratio of 80%: 20%. CNN model used in classification of digital images of herbs and spices is a model with 2 convolutional layers, where the first convolutional layer has 10 filters and the second convolutional layer has 20 filters. Each filter has a kernel matrix with a size of 3x3. The filter size at the pooling layer is 3x3 and the number of neurons in the hidden layer is 10. The activation function at the convolutional layer and hidden layer is tanh, and the activation function at the output layer is softmax. In this model, the accuracy of training data is 0.9875 and the loss value is 0.0769. The accuracy of testing data is 0.85 and the loss value is 0.4773. Meanwhile, testing new data with 3 images for each category produces an accuracy of 88.89%. Keywords: image classification, herbs and spices, CNN. 

scholarly journals Connected Component and Morphology Based Extraction of Arterial Centerlines of the Heart (CocomoBeach)

2008 ◽  
Author(s):  
Pieter Kitslaar ◽  
Michel Frenay ◽  
Elco Oost ◽  
Jouke Dijkstra ◽  
Berend Stoel ◽  
...  
Keyword(s):  
Coronary Artery ◽  
Training Data ◽  
Data Sets ◽  
Fast Marching ◽  
Geometrical Properties ◽  
Testing Data ◽  
Heart Region ◽  
Extraction Scheme ◽  
The Right ◽  
Coronary Artery Tree

This document describes a novel scheme for the automated extraction of the central lumen lines of coronary arteries from computed tomography angiography (CTA) data. The scheme first obtains a seg- mentation of the whole coronary tree and subsequently extracts the centerlines from this segmentation. The first steps of the segmentation algorithm consist of the detection of the aorta and the entire heart region. Next, candidate coronary artery components are detected in the heart region after the masking of the cardiac blood pools. Based on their location and geometrical properties the structures representing the right and left arterties are selected from the candidate list. Starting from the aorta, connections between these structures are made resulting in a final segmentation of the whole coronary artery tree, A fast-marching level set method combined with a backtracking algorithm is employed to obtain the initial centerlines within this segmentation. For all vessels a curved multiplanar reformatted image (CMPR) is constructed and used to detect the lumen contours. The final centerline was then defined by determining the center of gravity of the detected lumen in the transversal CMPR slices. Within the scope of the MICCAI Challenge “Coronary Artery Tracking 2008”, the coronary tree segmentation and centerline extraction scheme was used to automatically detect a set of centerlines in 24 datasets. For 8 data sets reference centerlines were available. This training data was used during the development and tuning of the algorithm. Sixteen other data sets were provided as testing data. Evaluation of the proposed methodology was performed through submission of the resulting centerlines to the MICCAI Challenge website

scholarly journals Classification of Bird Based on Face Types Using Gray Level Co-Occurrence Matrix (GLCM) Feature Extraction Based on the k-Nearest Neighbor (K-NN) Algorithm

2021 ◽  
Vol 6 (2) ◽  
pp. 111-119
Author(s):  
Daurat Sinaga ◽  
Feri Agustina ◽  
Noor Ageng Setiyanto ◽  
Suprayogi Suprayogi ◽  
Cahaya Jatmoko
Keyword(s):  
Feature Extraction ◽  
Nearest Neighbor ◽  
Correlation Energy ◽  
Training Data ◽  
Gray Level ◽  
K Nearest Neighbor ◽  
Testing Data ◽  
Occurrence Matrix

Indonesia is one of the countries with a large number of fauna wealth. Various types of fauna that exist are scattered throughout Indonesia. One type of fauna that is owned is a type of bird animal. Birds are often bred as pets because of their characteristic facial voice and body features. In this study, using the Gray Level Co-Occurrence Matrix (GLCM) based on the k-Nearest Neighbor (K-NN) algorithm. The data used in this study were 66 images which were divided into two, namely 55 training data and 11 testing data. The calculation of the feature value used in this study is based on the value of the GLCM feature extraction such as: contrast, correlation, energy, homogeneity and entropy which will later be calculated using the k-Nearest Neighbor (K-NN) algorithm and Eucliden Distance. From the results of the classification process using k-Nearest Neighbor (K-NN), it is found that the highest accuracy results lie at the value of K = 1 and at an degree of 0 ° of 54.54%.

scholarly journals Penerapan Artificial Neural Network (ANN) untuk Memprediksi Perubahan Derajat Miopia pada Manusia

2020 ◽  
Vol 10 (1) ◽  
pp. 53
Author(s):  
Ni Kadek Emik Sapitri ◽  
I Putu Eka N. Kencana ◽  
Luh Putu Ida Harini
Keyword(s):  
Training Data ◽  
Fuzzy Membership ◽  
Fuzzy Membership Function ◽  
Individual Identity ◽  
Vision Disorder ◽  
Testing Data ◽  
Visual Disorders ◽  
Artificial Neural Network Ann ◽  
The Right ◽  
And Behavior

Myopia is a vision disorder that causes the sufferers unable to see distant objects. The degree of myopia in humans can changes, both increasing and decreasing. The increasing of myopia degree is proportional to the potential of other visual disorders, such as cataracts, retinal detachment, and glaucoma. Therefore, the increasing of myopia degree needs to be watched out. Several previous studies only considered the time factor in predicting the changes of myopia degree. In fact, the changes of myopia degree also influenced by some factors that related to individual identity and behavior. This study aims to predict the changes of myopia degree in humans based on some factors that causes myopia.. This study uses data that has been scaled with the fuzzy membership function to be processed with ANN for predicting the changes of myopia degree. By ANN 6-2-3 architecture that uses 80 training data, 20 testing data, and 1 predictive data, the prediction result of the changes of  myopia degree in the right eye is 1.1 dioptri, in the left eye is 1.2 dioptri and the accumulated of both is 2.3 dioptri with accuration values 87.79%, 78.47%, and 83.21%.

scholarly journals Evaluasi Ekstraksi Fitur GLCM dan LBP Menggunakan Multikernel SVM untuk Klasifikasi Batik

2021 ◽  
Vol 5 (1) ◽  
pp. 1-9
Author(s):  
Pulung Nurtantio Andono ◽  
Eko Hari Rachmawanto
Keyword(s):  
Local Binary Pattern ◽  
Training Data ◽  
Polynomial Kernel ◽  
Support Vector ◽  
Gaussian Kernels ◽  
Central Java ◽  
Testing Data ◽  
Different Color ◽  
Occurrence Matrix

Batik as one of Indonesia's cultural heritages has various types, motifs and colors. A batik may have almost the same motif with a different color or vice versa, therefore it requires a classification of batik motifs. In this study, a printed batik was used with various coastal batik motifs in Central Java. The algorithm for classification is selected Support Vector Machine (SVM) with feature extraction of the Gray Level Co-Occurrence Matrix (GLCM) and Local Binary Pattern (LBP). SVM has the advantage of grouping data with small amounts and short operation times. GLCM as an extractive feature for recognizing batik textures and LBP was chosen to do spot pattern recognition. In the experiment, we have used 160 images of batik motifs which are divided into two, namely 128 training data and 32 testing data. The accuracy results obtained from the SVM, GLCM and LBP algorithms produce 100% accuracy in polyniomial, linear and gaussian kernels with distances at GLCM 1, 3, and 5, where at a distance of 1 linear kernel is 78.1%, gaussian 93.7%. At a distance of 3 linear kernels 75%, gaussian 87.5% and at a distance of 5 linear kernels 84.3%, gaussian 87.5%. In the SVM and GLCM algorithms the resulting accuracy is at a distance of 1 with a polynomial kernel 96.8%, linear 68.7%, and gaussian 75%. At distance 3, the polynomial kernel is 100%, linear 71.8%, and gaussian 78.1%, while for distance 5, the polynomial kernel is 87.5%, linear 75%, and gaussian 81.2%.

scholarly journals South German Credit Data Classification Using Random Forest Algorithm to Predict Bank Credit Receipts

2020 ◽  
Vol 3 (2) ◽  
Author(s):  
Yoga Religia ◽  
Gatot Tri Pranoto ◽  
Egar Dika Santosa
Keyword(s):  
Data Mining ◽  
Random Forest ◽  
Large Data ◽  
Classification Algorithm ◽  
Large Data Sets ◽  
Training Data ◽  
Data Sets ◽  
Testing Data ◽  
Credit Data

Normally, most of the bank's wealth is obtained from providing credit loans so that a marketing bank must be able to reduce the risk of non-performing credit loans. The risk of providing loans can be minimized by studying patterns from existing lending data. One technique that can be used to solve this problem is to use data mining techniques. Data mining makes it possible to find hidden information from large data sets by way of classification. The Random Forest (RF) algorithm is a classification algorithm that can be used to deal with data imbalancing problems. The purpose of this study is to discuss the use of the RF algorithm for classification of South German Credit data. This research is needed because currently there is no previous research that applies the RF algorithm to classify South German Credit data specifically. Based on the tests that have been done, the optimal performance of the classification algorithm RF on South German Credit data is the comparison of training data of 85% and testing data of 15% with an accuracy of 78.33%.

Sign in / Sign up

Export Citation Format

Share Document