Deep Belief Network Method Research Articles

Sentiment Analysis on Twitter Using Deep Belief Network Optimized with Particle Swarm Optimization

Deep Belief Network is a type of artificial neural network that is widely used in machine learning and deep learning tasks that allows it to learn hierarchical representations of the input data. However, Deep Belief Network has a drawback of being sensitive to hyperparameters. DBN has several hyperparameters that need to be chosen appropriately for the network to function effectively. Poor hyperparameter choices can lead to unstable training or poor performance. Therefore, the Particle Swarm Optimization algorithm is used to search for the best hyperparameters, which can lead to stable training and improved performance. The purpose of this study is to analyze public sentiment on Twitter using the Deep Belief Network method and to optimize it using Particle Swarm Optimization. The evaluation results obtained in this study are 71.4% accuracy, 71.7% precision, 71.4% recall, and 71.2% F1-score in the Deep Belief Network model which is optimized by the Particle Swarm Optimization algorithm, whereas when compared with the Deep Belief Network model alone gets evaluation results of 68.3% Accuracy, 69.0% Precision, 68.3% Recall and 68.0% F1-score. These results indicate that the use of the Particle Swarm Optimization algorithm is quite influential in analyzing sentiment in text.

Multi-Objective Seagull Optimization Algorithm with Deep Learning-Enabled Vulnerability Detection for Secure Cloud Environments

Cloud computing (CC) is an internet-enabled environment that provides computing services such as networking, databases, and servers to clients and organizations in a cost-effective manner. Despite the benefits rendered by CC, its security remains a prominent concern to overcome. An intrusion detection system (IDS) is generally used to detect both normal and anomalous behavior in networks. The design of IDS using a machine learning (ML) technique comprises a series of methods that can learn patterns from data and forecast the outcomes consequently. In this background, the current study designs a novel multi-objective seagull optimization algorithm with a deep learning-enabled vulnerability detection (MOSOA-DLVD) technique to secure the cloud platform. The MOSOA-DLVD technique uses the feature selection (FS) method and hyperparameter tuning strategy to identify the presence of vulnerabilities or attacks in the cloud infrastructure. Primarily, the FS method is implemented using the MOSOA technique. Furthermore, the MOSOA-DLVD technique uses a deep belief network (DBN) method for intrusion detection and its classification. In order to improve the detection outcomes of the DBN algorithm, the sooty tern optimization algorithm (STOA) is applied for the hyperparameter tuning process. The performance of the proposed MOSOA-DLVD system was validated with extensive simulations upon a benchmark IDS dataset. The improved intrusion detection results of the MOSOA-DLVD approach with a maximum accuracy of 99.34% establish the proficiency of the model compared with recent methods.

Sentiment Analysis for Online Product Reviews and Recommendation Using Deep Learning Based Optimization Algorithm

Recently, online shopping is becoming a popular means for users to buy and consume with the advances in Internet technologies. Satisfaction of users could be efficiently improvised by carrying out a Sentiment Analysis (SA) of larger amount of user reviews on e-commerce platform. But still, it is a challenge to envision the precise sentiment polarity of the user reviews due to the modifications in sequence length, complicated logic, and textual order. In this study, we propose a Hybrid-Flash Butterfly Optimization with Deep Learning based Sentiment Analysis (HFBO-DLSA) for Online Product Reviews. The presented HFBO-DLSA technique mainly aims to determine the nature of sentiments based on online product reviews. For accomplishing this, the presented HFBO-DLSA technique applies data pre-processing at the preliminary stage to make it compatible. Besides, the HFBO-DLSA model uses deep belief network (DBN) model for classification. The HFBO algorithm is used as a hyperparameter tuning process to improve the SA performance of the DBN method. The experimental validation of the presented HFBO-DLSA method has been tested under a set of datasets. The experimental results reveal that the HFBO-DLSA approach surpasses recent techniques in terms of SA outcomes. Specifically, when compared to various existing models on the Canon dataset, the HFBO-DLSA technique achieves remarkable results with an accuracy of 97.66%, precision of 98.54%, recall of 94.64%, and an F-score of 96.43%. In comparative analysis, other approaches such as ACO, SVM, and NN exhibit poorer performance, while TextCNN, BiLSTM, and RCNN approaches yield slightly improved SA results.

Optimized deep belief network and unsupervised deep learning methods for disease prediction

Due to the vast amount of patient health data, automated healthcare systems still struggle to classify and diagnose various ailments. Learning redundant data also reduces categorization accuracy. A Deep Belief Network (DBN) has been used to precisely extract the most important aspects from clinical data by ignoring irrelevant/redundant features. Due of many learning variables, training is complicated. Similarly, the hybrid model has been employed by ensemble Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM) and Support Vector Machine (SVM) to categorize diseases. But, its efficiency depends on the proper choice of kernels and hyper-parameters. Therefore, this paper develops an efficient feature extraction and classification model for healthcare systems. First, several medical data related to the patient’s health are collected. Then, an Optimized DBN (ODBN) model is presented for maximizing the accurateness of DBN by optimizing the learning variables depends on the Ant Lion Optimization (ALO) algorithm. With learning ODBN, the most relevant features are extracted with reduced computational complexity. After that, the CNN-LSTM with Unsupervised Fine-tuned Deep Self-Organizing Map (UFDSOM)-based classifier model is designed to categorize the extracted features into categories of illnesses. In this novel classifier, dropout normalization and parameter tuning processes are applied to avoid overfitting and optimize the hyper-parameters, which results in a less training period. In the end, studies utilizing publically accessible datasets show that the ODBN with CNN-LSTM-UFDSOM system outperforms classical models by 98.23%.

Fault Diagnosis of Bearings Using Wavelet Packet Energy Spectrum and SSA-DBN

To enhance fault characteristics and improve fault detection accuracy in bearing vibration signals, this paper proposes a fault diagnosis method using a wavelet packet energy spectrum and an improved deep confidence network. Firstly, a wavelet packet transform decomposes the original vibration signal into different frequency bands, fully preserving the original signal’s frequency information, and constructs feature vectors by extracting the energy of sub-frequency bands via the energy spectrum to extract and enhance fault feature information. Secondly, to minimize the time-consuming manual parameter adjustment procedure and increase the diagnostic accuracy, the sparrow search algorithm–deep belief network method is proposed, which utilizes the sparrow search algorithm to optimize the hyperparameters of the deep belief networks and reduce the classification error rate. Finally, to verify the effectiveness of the method, the rolling bearing data from Casey Reserve University were selected for verification, and compared to other commonly used algorithms, the proposed method achieved 100% and 99.34% accuracy in two sets of comparative experiments. The experimental results demonstrate that this method has a high diagnostic rate and stability.

Analysis and Countermeasures of Psychological Characteristics in College Students’ Psychological Education Based on Artificial Intelligence

ABSTRACT It is true that with society entering a new era, the pressure of competition for employment of college students is increasing, and society has higher requirements for the quality of high-tech talents, which can result in mental health issues for college students. Therefore, focusing on psychological education for college students and improving their psychological quality is essential. The analysis and countermeasure research on psychological characteristics has become crucial in this context. To address this issue, this work proposes an Improved Particle Swarm Optimization (IPSO) Deep Belief Network (DBN) method to analyze the psychological characteristics of college students. The proposed method first uses DBN to analyze the psychological characteristics of college students. Secondly, the learning factors and inertia weights in PSO are dynamically adjusted to construct IPSO to address the problem of slow overall convergence speed caused by the initial randomization of weights in the DBN training process. IPSO is then combined with DBN to adjust the initial weights for DBN training. This work also proposes a series of countermeasures from different aspects to strengthen mental health education for college students. The validity of IPSO-DBN for college students’ psychological characteristics analysis is verified through comprehensive and systematic experiments, and the feasibility of college students’ mental health education countermeasures is confirmed via comparison. Overall, this work provides an innovative and effective approach for analyzing college students’ psychological characteristics and developing countermeasures to improve their mental health. It shows the potential of combining artificial intelligence with psychological education and highlights the importance of strengthening psychological health education for college students.

Henry Gas Solubility Optimization Algorithm based Feature Extraction in Dermoscopic Images Analysis of Skin Cancer.

Artificial Intelligence (AI) techniques have changed the general perceptions about medical diagnostics, especially after the introduction and development of Convolutional Neural Networks (CNN) and advanced Deep Learning (DL) and Machine Learning (ML) approaches. In general, dermatologists visually inspect the images and assess the morphological variables such as borders, colors, and shapes to diagnose the disease. In this background, AI techniques make use of algorithms and computer systems to mimic the cognitive functions of the human brain and assist clinicians and researchers. In recent years, AI has been applied extensively in the domain of dermatology, especially for the detection and classification of skin cancer and other general skin diseases. In this research article, the authors propose an Optimal Multi-Attention Fusion Convolutional Neural Network-based Skin Cancer Diagnosis (MAFCNN-SCD) technique for the detection of skin cancer in dermoscopic images. The primary aim of the proposed MAFCNN-SCD technique is to classify skin cancer on dermoscopic images. In the presented MAFCNN-SCD technique, the data pre-processing is performed at the initial stage. Next, the MAFNet method is applied as a feature extractor with Henry Gas Solubility Optimization (HGSO) algorithm as a hyperparameter optimizer. Finally, the Deep Belief Network (DBN) method is exploited for the detection and classification of skin cancer. A sequence of simulations was conducted to establish the superior performance of the proposed MAFCNN-SCD approach. The comprehensive comparative analysis outcomes confirmed the supreme performance of the proposed MAFCNN-SCD technique over other methodologies.

Knowledge-extracted deep learning diagnosis and its cloud-based management for multiple faults of chiller

Machine learning has the inspiring potential for fault prediction in HVAC systems, which is essential for the system energy efficiency. For the scenario of multiple faults, however, machine learning model is not efficient because the training datasets of this scenario are difficult to collect in the real applications. This paper proposes the novel knowledge-embedded deep belief network (DBN) method to diagnose the electronic-thermal and thermal-thermal multiple faults for chillers in the buildings. Firstly, the characteristics of electronic-thermal and thermal-thermal faults are analyzed through the experiments. Through detecting the sensor biases, the sensor-thermal faults are decoupled successfully. Secondly, the representative features are extracted using the proposed DBN model. In the fine-tuning process, the original DBN network is optimized through embedding the extracted knowledge rules. We integrate the knowledge-embedded DBN, extreme learning machine (ELM) and k-nearest neighbor (KNN) as the diagnosis model. The experimental datasets are used to validate the proposed multi-fault diagnosis method. The results show that its diagnosis performance is satisfactory when the training dataset of some fault is absent. Finally, we develop the cloud-based diagnosis and online management platform. The proposed method is deployed on the cloud to realize online diagnosis and smart management.

Automated climate prediction using pelican optimization based hybrid deep belief network for Smart Agriculture

A significant part of the economies of several nations throughout the world, including India, where agriculture accounts for 16% of the total economy. Due to their dynamic and turbulent natures, climate prediction in this region is the most difficult because statistical methodologies can't provide accurate predictions. Typically, weather forecasts were made using extraordinarily complex physics techniques that took into account various atmospheric conditions over a long time. These circumstances were usually unstable as a result of weather system disturbances, which led to the development of techniques for making unreliable predictions. In this study, an Automated Climate Prediction for Smart Agriculture is developed utilizing a Pelican Optimization-based Hybrid Deep Belief Network (ACP-POHDBN). The purpose of the ACP-POHDBN approach is to identify the appropriate meteorological conditions. The acp-POHDBN method uses the test data to calculate the appropriate weather conditions. Pre-processing, prediction, and hyperparameter tuning are the three key steps that make up the proposed ACP-POHDBN approach. The min-max normalization procedure is the primary procedure used by the ACP-POHDBN model to convert the meteorological data into a standard format. In addition to pre-processing data, the Deep Belief Network (DBN) model is used to forecast weather conditions. Finally, the DBN method's hyperparameters (epoch, batch size, and learning rate) may be properly tuned using the POA-based hyperparameter tuning technique. The development of a Pelican Optimization Algorithm (POA)-based hyperparameter optimizer for the process of predicting the climate demonstrates the originality of the study. The proposed algorithm over other recent approaches with maximum accuracy of 95.03%, sensitivity of 95.03%, specificity of 95.03%, and F-score of 95.03%.

Anas platyrhynchos optimizer with deep transfer learning-based gastric cancer classification on endoscopic images

<abstract><p>Gastric Cancer (GC) has been identified as the world's fifth most general tumor. So, it is important to diagnose the GC at initial stages itself to save the lives. Histopathological analysis remains the gold standard for accurate diagnosis of the disease. Though Computer-Aided Diagnostic approaches are prevalently applied in recent years for the diagnosis of diseases, it is challenging to apply in this case, due to the lack of accessible gastric histopathological image databases. With a rapid progression in the Computer Vision (CV) technologies, particularly, the emergence of medicinal image classifiers, it has become feasible to examine all the types of electron micrographs in a rapid and an effective manner. Therefore, the current research article presents an Anas Platyrhynchos Optimizer with Deep Learning-based Gastric Cancer Classification (APODL-GCC) method for the classification of GC using the endoscopic images. The aim of the proposed APODL-GCC method is to identify the presence of GC with the help of CV and Deep Learning concepts. Primarily, the APODL-GCC technique employs a contrast enhancement technique. Next, the feature extraction process is performed using a neural architectural search network model to generate a collection of feature vectors. For hyperparameter optimization, the Anas Platyrhynchos Optimizer (APO) algorithm is used which enhances the classification performance. Finally, the GC classification process is performed using the Deep Belief Network method. The proposed APODL-GCC technique was simulated using medical images and the experimental results established that the APODL-GCC technique accomplishes enhanced performance over other models.</p></abstract>

A Novel Framework for DDoS Attacks Detection Using Hybrid LSTM Techniques

The recent development of cloud computing offers various services on demand for organization and individual users, such as storage, shared computing space, networking, etc. Although Cloud Computing provides various advantages for users, it remains vulnerable to many types of attacks that attract cyber criminals. Distributed Denial of Service (DDoS) is the most common type of attack on cloud computing. Consequently, Cloud computing professionals and security experts have focused on the growth of preventive processes towards DDoS attacks. Since DDoS attacks have become increasingly widespread, it becomes difficult for some DDoS attack methods based on individual network flow features to distinguish various types of DDoS attacks. Further, the monitoring pattern of traffic changes and accurate detection of DDoS attacks are most important and urgent. In this research work, DDoS attack detection methods based on deep belief network feature extraction and Hybrid Long Short-Term Memory (LSTM) model have been proposed with NSL-KDD dataset. In Hybrid LSTM method, the Particle Swarm Optimization (PSO) technique, which is combined to optimize the weights of the LSTM neural network, reduces the prediction error. This deep belief network method is used to extract the features of IP packets, and it identifies DDoS attacks based on PSO-LSTM model. Moreover, it accurately predicts normal network traffic and detects anomalies resulting from DDoS attacks. The proposed PSO-LSTM architecture outperforms the classification techniques including standard Support Vector Machine (SVM) and LSTM in terms of attack detection performance along with the results of the measurement of accuracy, recall, f-measure, precision.

Design of Soft-Sensing Model for Alumina Concentration Based on Improved Deep Belief Network

Alumina concentration is an important parameter in the production process of aluminum electrolysis. Due to the complex production environment in the industrial field and the complex physical and chemical reactions in the aluminum reduction cell, nowadays it is still unable to carry out online measurement and real-time monitoring. For solving this problem, a soft-sensing model of alumina concentration based on a deep belief network (DBN) is proposed. However, the soft-sensing model may have some limitations for different cells and different periodic working conditions such as local anode effect, pole changing, and bus lifting in the same cell. The empirical mode decomposition (EMD) and particle swarm optimization (PSO) with the DBN are combined, and an EMD–PSO–DBN method that can denoize and optimize the model structure is proposed. The simulation results show that the improved soft-sensing model improves the accuracy and universality of prediction.

Automatic detection of sleep apnea from single-lead ECG signal using enhanced-deep belief network model

Sleep apnea (SLA) is a commonly reported sleep disorder that is characterized by frequent breathing interruptions during sleep. In recent years, various approaches have been made to developing a less complex and cost-effective process for diagnosing SLA patients, as opposed to using the inconvenient, complex and costly polysomnography test. This study proposed a novel approach of cascading two different types of Restricted Boltzmann Machine (RBM) in Deep Belief Networks (DBN) method for the SLA classification using single-lead electrocardiogram (ECG) signals. The proposed framework uses two types of RBM, namely Gaussian-Bernoulli, and Bernoulli-Bernoulli, which are modified forms of the Boltzmann Machine, to develop an enhanced- DBN (E-DBN) structure for significant feature learning and detection of SLA and normal events. At each ECG data signal, Heart Rate Variability (HRV) and ECG-Derived Respiration (EDR) signals are obtained from the 1-minute segmented ECG signal. The E-DBN model can also enhance the parameter detection performance as a keyfeature extractor and classification method. The Apnea-ECG datasets from physionet were used to train the presented fine-tuned E-DBN model and validate its performance for detecting SLA incidents. The proposed method shows significant improvements in performance when compared to other SLA detection methods for per-segment detection and achieved the highest accuracy of 89.11%, with specificity, sensitivity, and F1-score of 92.28%, 83.89%, and 0.913, respectively. For per-recording detection, accuracy of 97.17% and correction coefficient values of 0.938 are obtained. The proposed approach develops a method that analyses single-lead electrocardiography data from patients and diagnoses the SLA condition of patients.

Islanding power quality detection using lighting search optimization with deep learning model on distributed generation systems

Recently, the significant penetration of distributed generations (DGs) in distribution networks has made new challenging issues in network protection. A major challenge exists in the distribution networks in the identification of islanding conditions, where a portion of the distribution network with DGs is accidentally detached from the central grid due to protection relay and circuit breaker operation. Since it results in damaging the DGs and their equipment, effective islanding condition detection model becomes essential. Besides, non-detected islanding conditions can lead to voltage and frequency deviances from the normal ranges, inappropriate operation of protection, and personnel hazards. Therefore, the recent developments in deep learning (DL) can be applied to perform island power quality detection process. In this view, this article introduces an Islanding Power Quality Detection using Lighting Search Optimization with Deep Learning (IPQD-LSODL) model. The proposed IPQD-LSODL model mainly aims to find the events in islanding power quality (IPQ) and non-islanding power quality (NIPQ). The proposed model initially applies down-sampling empirical mode decomposition (DEMD) approach which effectively filters out the basic signal from the polluted ones. In addition, deep belief network (DBN) model is used to classify the events into IPQ and NIPQ. Moreover, the hyperparameters of the DBN method were optimally chosen by the use of LSO algorithm with an intention of accomplishing maximum classification accuracy. The performance validation of the IPQD-LSODL model is carried out and validated through comparison study with existing models. The IPQD-LSODL method achieved an accuracy of 99.91 percent in the class 5. The results implied the promising performances of the IPQD-LSODL method over recent approaches.

Identification of Encrypted Traffic Using Advanced Mathematical Modeling and Computational Intelligence

This paper proposed a hybrid approach for the identification of encrypted traffic based on advanced mathematical modeling and computational intelligence. Network traffic identification is the premise and foundation of improving network management, service quality, and application security. It is also the focus of network behavior analysis, network planning and construction, network anomaly detection, and network traffic model research. With the increase in user and service requirements, many applications use encryption algorithms to encrypt traffic during data transmission. As a result, traditional traffic classification methods classify encrypted traffic on the network, which brings great difficulties and challenges to network monitoring and data mining. In our article, a nonlinear modified DBN method is proposed and applied to encrypted traffic identification. Firstly, based on Deep Belief Networks (DBN), this paper introduces the proposed Eodified Elliott (ME)-DBN model, analyzes the function image, and presents the ME-DBN learning algorithm. Secondly, this article designs an encrypted traffic recognition model based on the ME-DBN model. Feature extraction is carried out by training the ME-DBN model, and finally, classification and recognition are carried out by the classifier. The experimental results on the ISCX VPN-non-VPN database show that the MEDBN method proposed in this article can enhance the classification and recognition rate and has better robustness to encrypt traffic recognition from different software.

An Optimized Deep Belief Network for Land Cover Classification Using Synthetic-Aperture Radar Images and Landsat Images

Abstract This paper intends to propose an automated deep learning-based land cover classification model of remote sensing images. The model includes (i) pre-processing, (ii) feature extraction and (iii) classification. The captured synthetic-aperture radar (SAR) and Landsat-8 images are initially pre-processed using the Gabor filtering model. Subsequently, from SAR images the gray-level-co-occurrence matrix-based texture characteristics are extracted, and temperature vegetation index-based characteristics, normalized vegetation index-based features, normalized difference index-based features and coloration index features are extracted from Landsat-8 images. Finally, the extracted features are subjected to an optimized deep belief network (DBN), where the weight is fine-tuned by the optimization logic. For this, a new Sunflower adopted Red Deer (SARD) algorithm is introduced in this work that hybrids the concept of Red Deer algorithm and Sunflower optimization. The performance of the proposed classification model is compared over other conventional models concerning different measures. Especially, the accuracy of the presented work (SARD+DBN) for Testcase3 is 5, 7, 6 and 30% better than existing DA + DBN, JA + DBN, SLnO+DBN and LA + DBN methods, respectively.

Short-Term Traffic Flow Prediction with Structure-Optimized Deep Belief Network

This paper proposes a structure-optimized deep belief network method for short-term traffic flow forecast, which is used to solve the problems of too simple training data in deep learning short-term traffic flow forecast and random selection of model structure construction parameters. We constructed a deep belief network short-term traffic flow forecast model that can simultaneously train three types of traffic data related to the predicted node traffic volume, enhance the spatiotemporal correlation of predictions, and overcome the shortcomings of too single training data. At the same time, we optimize the short-term traffic flow prediction model structure of the deep belief network; and use the T-PSO algorithm to optimize the hidden layer structure parameters of the prediction model. It can avoid the decrease in the practicability of the model caused by the random selection of the model structure construction parameters. The experimental results show that the method of structure-optimized deep belief network is feasible and effective, and the prediction accuracy is better than the classic deep learning prediction model.

Diagnosis of COVID-19 patients by adapting hyper parametertuned deep belief network using hosted cuckoo optimization algorithm

ABSTRACT COVID-19 is an infection caused by recently discovered corona virus. The symptoms of COVID-19 are fever, cough and dumpiness of breathing. A quick and accurate identification is essential for an efficient fight against COVID-19. A machine learning technique is initiated for categorizing the chest x-ray images into two cases: COVID-19 positive case or negative case. In this manuscript, the categorization of COVID-19 can be determined by hyper parameter tuned deep belief network using hosted cuckoo optimization algorithm. At first, the input chest x-ray images are pre-processed for removing noises. In this manuscript, the deep belief network method is enhanced by hosted cuckoo optimization approach for getting optimum hyper tuning parameters. By this, exact categorization of COVID-19 is attained effectively. The proposed methodology is stimulated at MATLAB. The proposed approach attains 28.3% and 23.5% higher accuracy for Normal and 32.3% and 31.5% higher accuracy for COVID-19, 19.3% and 28.5% higher precision for Normal and 45.3% and 28.5% higher precision for COVID-19, 20.3% and 21.5% higher F-score for Normal and 40.3% and 21.5% higher F-score for COVID-19. The proposed methodology is analyzed using two existing methodologies, as Convolutional Neural Network with Social Mimic Optimization (CNN-SMO) and Support Vector Machine classifier using Bayesian Optimization algorithm (SVM-BOA).

Intelligent Fault Severity Detection of Rotating Machines Based on VMD-WVD and Parameter-Optimized DBN

An intelligent fault severity detection method based on variational mode decomposition- (VMD-) Wigner-Ville distribution (WVD) and sparrow search algorithm- (SSA-) optimized deep belief network (DBN) is suggested to address the problem that typical fault diagnostic algorithms are inappropriate for extremely comparable vibration signals when the samples are insufficient. VMD is used to process the original vibration signal to obtain the band intrinsic mode functions (BIMFs) with different frequencies. WVD produces the two-dimensional spectrum of the key BIMF with the highest variance contribution rate. The input sample of DBN is composed of a characteristic matrix formed by the two-dimensional spectrum of multiple fault signals. DBN’s learning rate and batch size are both tuned globally by SSA, which has a significant influence on network error. The fitness function in the parameter optimization process is the network’s root mean square error (RMSE). Finally, the input samples are loaded into a DBN that has the best structure for detecting severity. Experiments show that, based on VMD-WVD and SSA-DBN, accuracy of the fault severity detection model for rotating machines, which has good generalization ability and robustness, can reach 98%. Compared with BPNN, the traditional DBN, VMD-DBN, VMD-PSO-DBN, and other methods, the proposed algorithm has strong adaptive feature extraction ability and generalization of application.

Class metric regularized deep belief network with sparse representation for fault diagnosis

This paper proposes a joint class metric and sparse representation regularized deep belief network (J-DBN) method for intelligent fault diagnosis of the rotary equipment. In this novel method, the joint class metric and sparse representation regularized DBN is considered as a pretraining method to extract data features. It combines advantages of both class metric and sparse representation, which can optimize the distance of features in the same class and penalize the distance of features in different classes, and generate sparse features. Specifically, a new metric matrix is constructed to avoid using the same structural parameters for the local structure of each sample. The J-DBN-based fault diagnosis is implemented by the pretraining learning method, which contributes to better classification capabilities. Finally, gearbox and bearing fault diagnosis experiments are conducted to validate the effectiveness and the superiority of the proposed method. The results show that the ability of the J-DBN method to extract features is significantly enhanced, and the clustering of features of the same data is more obvious; furthermore, the proposed method has higher diagnostic accuracy than other fault diagnosis methods.