A convolutional deep belief network for polarimetric SAR data feature extraction

2016 ◽  
Author(s):  
Radu Tanase ◽  
Mihai Datcu ◽  
Dan Raducanu
Keyword(s):  
Feature Extraction ◽  
Deep Belief Network ◽  
Polarimetric Sar ◽  
Belief Network ◽  
Sar Data

Deep Appearance Model and Crow-Sine Cosine Algorithm-Based Deep Belief Network for Age Estimation

2021 ◽  
Vol 12 (3) ◽  
pp. 185-207
Author(s):  
Anjali A. Shejul ◽  
Kinage K. S. ◽  
Eswara Reddy B.
Keyword(s):  
Feature Extraction ◽  
Age Estimation ◽  
Learning Algorithm ◽  
Directional Pattern ◽  
Input Image ◽  
Deep Belief Network ◽  
Appearance Model ◽  
Belief Network ◽  
Sine Cosine Algorithm ◽  
New Feature

Age estimation has been paid great attention in the field of intelligent surveillance, face recognition, biometrics, etc. In contrast to other facial variations, aging variation presents several unique characteristics, which make age estimation very challenging. The overall process of age estimation is performed using three important steps. In the first step, the pre-processing is performed from the input image based on Viola-Jones algorithm to detect the face region. In the second step, feature extraction is done based on three important features such as local transform directional pattern (LTDP), active appearance model (AAM), and the new feature, deep appearance model (Deep AM). After feature extraction, the classification is carried out based on the extracted features using deep belief network (DBN), where the DBN classifier is trained optimally using the proposed learning algorithm named as crow-sine cosine algorithm (CS).

scholarly journals Application of Multiacoustic Data in Feature Extraction of Anemometer

Complexity ◽  
2021 ◽  
Vol 2021 ◽  
pp. 1-10
Author(s):  
Dawei Chen ◽  
Xu Guo
Keyword(s):  
Feature Extraction ◽  
Wind Power ◽  
Evaluation Method ◽  
Training Model ◽  
Classification Algorithm ◽  
Deep Belief Network ◽  
Evaluation Function ◽  
Trust Network ◽  
Belief Network ◽  
Acoustic Data

The acoustic characteristics of wind instruments are a major feature in the field of vocal music. This paper studies the application effect of wind power instrument feature extraction based on multiacoustic data. Combined with the acoustic data training model, the classification algorithm based on deep trust network is used to process multiple acoustic data. Using multiple acoustic data for feature extraction, the recognition and matching between multiple acoustic data and wind measuring instrument are realized. The experiment not only evaluates the error of the network classification algorithm but also describes the evaluation function of the deep belief network classification algorithm in the system. The traditional SNR evaluation method is used to improve the deficiency of evaluation function. Through the deep belief network classification algorithm for self-learning, the instrument recognition method with strong applicability is established. Finally, the effectiveness of multiacoustic data in wind power instrument feature extraction is verified.

An optimization-based deep belief network for the detection of phishing e-mails

2020 ◽  
Vol 54 (4) ◽  
pp. 529-549
Author(s):  
Arshey M. ◽  
Angel Viji K. S.
Keyword(s):  
Feature Extraction ◽  
Feature Selection ◽  
Optimization Algorithm ◽  
Information Needs ◽  
Selection Process ◽  
Deep Belief Network ◽  
Data Set ◽  
Content Type ◽  
Belief Network ◽  
E Mail

PurposePhishing is a serious cybersecurity problem, which is widely available through multimedia, such as e-mail and Short Messaging Service (SMS) to collect the personal information of the individual. However, the rapid growth of the unsolicited and unwanted information needs to be addressed, raising the necessity of the technology to develop any effective anti-phishing methods.Design/methodology/approachThe primary intention of this research is to design and develop an approach for preventing phishing by proposing an optimization algorithm. The proposed approach involves four steps, namely preprocessing, feature extraction, feature selection and classification, for dealing with phishing e-mails. Initially, the input data set is subjected to the preprocessing, which removes stop words and stemming in the data and the preprocessed output is given to the feature extraction process. By extracting keyword frequency from the preprocessed, the important words are selected as the features. Then, the feature selection process is carried out using the Bhattacharya distance such that only the significant features that can aid the classification are selected. Using the selected features, the classification is done using the deep belief network (DBN) that is trained using the proposed fractional-earthworm optimization algorithm (EWA). The proposed fractional-EWA is designed by the integration of EWA and fractional calculus to determine the weights in the DBN optimally.FindingsThe accuracy of the methods, naive Bayes (NB), DBN, neural network (NN), EWA-DBN and fractional EWA-DBN is 0.5333, 0.5455, 0.5556, 0.5714 and 0.8571, respectively. The sensitivity of the methods, NB, DBN, NN, EWA-DBN and fractional EWA-DBN is 0.4558, 0.5631, 0.7035, 0.7045 and 0.8182, respectively. Likewise, the specificity of the methods, NB, DBN, NN, EWA-DBN and fractional EWA-DBN is 0.5052, 0.5631, 0.7028, 0.7040 and 0.8800, respectively. It is clear from the comparative table that the proposed method acquired the maximal accuracy, sensitivity and specificity compared with the existing methods.Originality/valueThe e-mail phishing detection is performed in this paper using the optimization-based deep learning networks. The e-mails include a number of unwanted messages that are to be detected in order to avoid the storage issues. The importance of the method is that the inclusion of the historical data in the detection process enhances the accuracy of detection.

scholarly journals HYPERSPECTRAL DATA FEATURE EXTRACTION USING DEEP BELIEF NETWORK

2016 ◽  
Vol 9 (4) ◽  
pp. 1991-2009 ◽  
Author(s):  
Jiang Xinhua ◽  
Xue Heru ◽  
Zhang Lina ◽  
Zhou Yanqing
Keyword(s):  
Feature Extraction ◽  
Hyperspectral Data ◽  
Deep Belief Network ◽  
Belief Network

scholarly journals Biometric Iris Recognition of Person from an Image at Long Distance using Chronological Monarch Butterfly Optimization based Deep Belief Network

2019 ◽  
Vol 8 (2S8) ◽  
pp. 1975-1983
Keyword(s):  
Feature Extraction ◽  
Hough Transform ◽  
Iris Recognition ◽  
Median Filter ◽  
Monarch Butterfly ◽  
Deep Belief Network ◽  
Long Distance ◽  
Belief Network ◽  
Monarch Butterfly Optimization

Now days, for the identification of personal information of a person, biometrics is mostly used. Also for the personal identification, the recognition of eye based biometric feature extraction is the most powerful tool. The biometric is an important identity to identify the individual. But in real time it is quite difficult to capture the better quality of iris images. The images obtained are more degraded due to the lack of texture, blur. In this paper, more convenient method is presented for extracting the features of biometric images. The method Iris Recognition at-a Distance (IAAD) is used to extract the iris features of biometric image and to enhance the quality of an image in a biometric system. The Chronological Monarch Butterfly Optimization -based Deep Belief Network (Chronological MBO-based DBN) is proposed for iris recognition to get better accuracy. The Monarch Butterfly Optimization algorithm is used to arrange the Chronological assumption of an iris image. Also, the Hough Transform algorithm is used for detection of iris circle and edge. The scaT T loop descriptor and the Local Gradient Pattern (LGP) are used for feature extraction, which is fed to the Chronological MBO-based DBN for iris recognition that enhances the accuracy. The Daugman’s rubber sheet model, median filter and trained neural network are used for normalization and segmentation. The UBIRIS.v1 database is used to take an iris recognition images and MATLAB is used for programming of for reading the iris images and for performing the Hough transform operations. The iris recognition at a distance 4 to 8 meter is done with the help of simulation result. The performance is analyzed based on the metrics, like False Acceptance Rate (FAR), accuracy, and False Rejection Rate (FRR) with the value of 0.4847%, 96.078%, and 0.4745%

scholarly journals Arrhythmia Detection using Deep Belief Network Extracted Features from ECG Signals

2021 ◽  
Vol 12 (6) ◽  
pp. 0-0
Keyword(s):  
Feature Extraction ◽  
Performance Enhancement ◽  
Model Performance ◽  
Deep Belief Network ◽  
Arrhythmia Detection ◽  
Belief Network ◽  
Ecg Signals ◽  
Increase In Accuracy ◽  
Simple Logistic ◽  
Electrocardiogram Ecg

Arrhythmia is a disorder of the heart caused by the erratic nature of heartbeats occurring due to conduction failures of the electrical signals in the cardiac muscle. In recent years, research galore has been done towards accurate categorization of heartbeats and electrocardiogram (ECG)-based heartbeat processing. Accurate categorization of different heartbeats is an important step for diagnosis of arrhythmia. This paper primarily focuses on effective feature extraction of the ECG signals for model performance enhancement using an unsupervised Deep Belief Network (DBN) pipelined onto a simple Logistic Regression (LR) classifier. We compare and evaluate the results of data feature enrichment against plain, non-enriched data based on the metrics of precision, recall, specificity, and F1-score and report the extent of increase in performance. Also, we compare the performance of the DBN-LR pipeline with a 1D convolution technique and find that the DBN-LR algorithm achieves a 5% and 10% increase in accuracy when compared to 1D convolution and no feature extraction using DBN respectively.

scholarly journals NEW APPROACH FOR ONLINE ARABIC MANUSCRIPT RECOGNITION BY DEEP BELIEF NETWORK

2018 ◽  
Vol 58 (5) ◽  
pp. 297
Author(s):  
Benbakreti Samir ◽  
Aoued Boukelif
Keyword(s):  
Neural Network ◽  
Feature Extraction ◽  
Deep Learning ◽  
Success Rate ◽  
Character Recognition ◽  
Recognition Rate ◽  
Deep Belief Network ◽  
New Approach ◽  
Belief Network ◽  
Network Methods

In this paper, we present a neural approach for an unconstrained Arabic manuscript recognition using the online writing signal rather than images. First, we build the database which contains 2800 characters and 4800 words collected from 20 different handwritings. Thereafter, we will perform the pretreatment, feature extraction and classification phases, respectively. The use of a classical neural network methods has been beneficial for the character recognition, but revealed some limitations for the recognition rate of Arabic words. To remedy this, we used a deep learning through the Deep Belief Network (DBN) that resulted in a 97.08% success rate of recognition for Arabic words.

scholarly journals Breast Cancer Detection Using Deep Belief Network by Applying Feature Extraction on Various Classifiers

2021 ◽  
Vol 12 (1S) ◽  
pp. 471-487
Author(s):  
Prasath Alias Surendhar S, Et. al.
Keyword(s):  
Breast Cancer ◽  
Feature Extraction ◽  
Deep Belief Network ◽  
Gradient Boosting ◽  
Scale Invariant ◽  
Belief Network ◽  
Cancer Breast ◽  
Tree Classifier ◽  
Breast Cells ◽  
Scale Invariant Feature

Among various types of cancer, breast cancer is considered as second largest hazardous diseases that cause death. Initially, small lump like structure grow from breast cells which are considered as malignant lumps. To find the available of malignancy in breast area, several checkups such as self-test and periodic has to be done to reduce the death rate due to breast cancer. But, the classification of breast cancer by medical physicians using available techniques is not sufficient, so it is important to improve the classification technique using neural network. The four important phases namely preprocessing, segmentation, feature extraction, and classification can be done in constructed Deep Belief Network (DBN) whereas preprocessing makes to remove the noise and artifacts of mammogram image and then the glands are enhanced. The preprocessed output is given to fuzzy c-means segmentation process with the help of masking. Again the feature extractors such as Scale Invariant Feature Transform (SIFT) and Speeded up Robust Transform (SURF) made to apply on two types of classifiers such as gradient boosting tree classifier and adaboost classifier. The examination is done interms of parameters such as accuracy, precision, recall and F1 score.

Sign in / Sign up

Export Citation Format

Share Document