Hybrid Deep Learning Method Research Articles

A Novel Hybrid Deep Learning Method for Accurate Exchange Rate Prediction

A Novel Hybrid Deep Learning Method for Accurate Exchange Rate Prediction

FlexLogNet: A flexible deep learning-based well-log completion method of adaptively using what you have to predict what you are missing

Well logs are essential tools for understanding the characteristics of subsurface formations and exploring petroleum resources. However, well logs are often missing randomly due to cost constraints, instrument failures, or other factors. Many methods have been developed for completing missing well logs, but these methods are all based on fixed types of known well-log inputs to predict specific types of missing logs. This fixed input–output mode severely limits the application of these methods in actual data, where the known and missing well-log types are often varying. To address this problem, we propose a hybrid deep learning method with two heads of heterogeneous graph neural network (HGNN) and fully connected network (FCN) to achieve mutual prediction among multiple types of well logs. It can adaptively use all known well logs to predict any missing well logs, achieving a very flexible and practical well log completion function of using what you have to complete what you are missing. Specifically, the HGNN head infers the inter-relationships among multiple well logs to predict normalized logs that contain detailed information, which achieved by using multiple independent kernels to extracting and aggregating the features of the multiple logs. The FCN head estimates the global statistics of the predicted logs, including means and standard deviations, for de-normalizing the well logs estimated by the HGNN head. Both the HGNN and FCN heads are trained simultaneously by a hybrid loss function to ensure the consistency of their predictions. Furthermore, we present an adaptive training strategy that leverages all well logs, including those with missing segments. We demonstrate the capability of our model using four well logs: gamma ray (GR), bulk density (RHOB), neutron porosity (NPHI), and compressional waves sonic (DTC). Theoretically, the model trained on other logs can also predict each other. Our approach yields high Pearson correlation coefficients and small root mean square error on a dataset obtained from an offshore North Sea field near Norway, demonstrating the efficacy of our proposed technique.

A ROBUST ENSEMBLE SEGMENTATION APPROACH AND OPPOSITION-BASED RAIN OPTIMIZATION ALGORITHM FOR ENHANCING ACUTE LYMPHOBLASTIC LEUKEMIA (ALL) DETECTION

Generally, leukemia is one of the blood cancers that can lead to death. A huge amount of immature WBCs in the bone marrow affect the healthy cells and this has become the leading cause of leukemia disease. Moreover, “Acute Lymphoblastic Leukemia (ALL)” is a type of blood cancer that is generally categorized with a huge amount of immature lymphocytes that are given as blast cells. But, the analysis of this model is dependent, boring as well as time-consuming based on the skills of hematologists. So, there is a requirement for the most desirable techniques to tackle these restrictions. Additionally, the accurate and automated diagnosis of ALL is a crucial task. In this case, an automated detection model of ALL using ensemble segmentation and a heuristic-assisted layer-improved hybrid deep learning approach is implemented. In the initial stage, the raw images are collected from benchmark datasets. Further, the image pre-processing is undergone by contrast enhancement and filtering process. After pre-processing, the image segmentation is carried out by ensemble segmentation, which is acquired by the region growing, K-medoids, and Fuzzy C-Means (FCM). From the segmented images, based on the mutual information, the pixel is selected optimally, in which the optimal pixel is identified by Opposition-based Rain Optimization Algorithm (OROA). Subsequently, the optimal pixels are fed as input to the Layer Improved Hybrid DenseNet and ResNet (LIH-DRNet) model that is constructed with DenseNet and ResNet, where some hyperparameters are tuned optimally by the improved ROA. At last, the performance is evaluated with diverse performance metrics. Thus, the findings reveal that the developed hybrid deep learning method achieves a higher detection rate to ensure the effectiveness of the model.

Mobile traffic prediction with attention-based hybrid deep learning

Mobile data consumption is ever-increasing due to the continuous emergence of bandwidth-hungry mobile applications as well as the drastic increase in number of mobile terminals resulted from the pervasive adoption of Internet of Things. In the 5G+ era, network operators have to deal with new technical challenges on network management, resource optimization and mobility scheduling, which require the ability to perform efficient mobile traffic prediction. Traditional prediction methods using convex optimization, Markov model and linear equations cannot describe the various nonlinearity changes of network traffic. In this paper, we propose a novel traffic prediction algorithm for mobile networks by introducing a hybrid deep learning method equipped with the attention mechanism, which analyzes the traffic in cellular zones to determine the spatiotemporal characteristics in the residential community. The hybrid deep learning model includes autoregressive integrated moving average mode (ARIMA) preprocessing with decomposition ability for time series, external spatial feature variables, sequence-to-sequence (seq2seq; attention-based CNN-LSTM) pretraining and extreme gradient boosting (XGBoost) optimization processing for integrating nonlinear relationships. The CNN method shows good nonlinear generalization ability to mine deep features of community traffic and capture the nonlinear trend. Our proposed approach first decomposes the community traffic data into linear and residual parts through ARIMA and then seizes the deep features of community traffic through the constructed pretraining seq2seq framework. Finally, the pretraining outcomes are further adjusted by the XGBoost method. Simulation results show that the proposed solution improves the accuracy of prediction apparently.

Sentiment classification in Hindi text using hybrid deep learning method

Sentiment classification in Hindi text using hybrid deep learning method

A novel hybrid deep CNN model for breast cancer classification using Lipschitz-based image augmentation and recursive feature elimination

Breast cancer is the leading cause of death for women worldwide. Despite its difficulties, the histopathological diagnosis must be used because modern imaging techniques cannot differentiate between benign and malignant tumors. This article introduces a hybrid deep learning method with recursive feature elimination for categorizing histopathology images of breast lesions as benign or malignant, using standard datasets, to increase diagnostic accuracy. The proposed four-stage model includes several novel techniques. First, a modified Lipschitz-based image augmentation technique is utilized to ensure a proportionate representation of various sample classes. K-fold cross-validation is employed to mitigate the bias that emerges from the selection of training and testing datasets. Then, for efficient feature extraction, a hybrid convolutional neural network (CNN) approach is used by integrating three CNNs (PResNet-34, FE-VGG-16, and M-AlexNet). With the proposed universum twin support vector machine-based recursive feature elimination technique, the initial 1,000 features extracted by each deep model are further reduced to 100. Finally, the deep features are subsequently combined into a set of 300 features, which are then fed into machine learning classifiers such as decision tree, k-nearest neighborhood, linear discriminant analysis, linear regression, and support vector machine. The results demonstrate that the proposed hybrid CNN model outperforms state-of-the-art techniques in accurately classifying breast lesions as benign or malignant, with an accuracy of 99.99 %, sensitivity of 99.56 %, specificity of 99.53 %, precision of 99.55 %, and F1 score of 99.77 %.

Diabetic retinopathy prediction based on vision transformer and modified capsule network

Diabetic retinopathy is considered one of the most common diseases that can lead to blindness in the working age, and the chance of developing it increases as long as a person suffers from diabetes. Protecting the sight of the patient or decelerating the evolution of this disease depends on its early detection as well as identifying the exact levels of this pathology, which is done manually by ophthalmologists. This manual process is very consuming in terms of the time and experience of an expert ophthalmologist, which makes developing an automated method to aid in the diagnosis of diabetic retinopathy an essential and urgent need. In this paper, we aim to propose a new hybrid deep learning method based on a fine-tuning vision transformer and a modified capsule network for automatic diabetic retinopathy severity level prediction. The proposed approach consists of a new range of computer vision operations, including the power law transformation technique and the contrast-limiting adaptive histogram equalization technique in the preprocessing step. While the classification step builds up on a fine-tuning vision transformer, a modified capsule network, and a classification model combined with a classification model, The effectiveness of our approach was evaluated using four datasets, including the APTOS, Messidor-2, DDR, and EyePACS datasets, for the task of severity levels of diabetic retinopathy. We have attained excellent test accuracy scores on the four datasets, respectively: 88.18%, 87.78%, 80.36%, and 78.64%. Comparing our results with the state-of-the-art, we reached a better performance.

Transfer learning-based hybrid deep learning method for gas-bearing distribution prediction with insufficient training samples and uncertainty analysis

The availability of sample data, required for seismicity-based reservoir prediction using deep learning (DL), is often limited by well availability. To alleviate this problem, this study proposes a hybrid DL method combining unsupervised learning, transfer learning (TL), and intelligent optimization algorithms to predict gas-bearing distributions. First, three unsupervised learning methods were used to optimize multicomponent seismic attribute data during preprocessing and improve the sample data quality. Subsequently, a convolutional neural network (CNN) model was pre-trained using synthetic data, and the adaptive mutation particle swarm optimization (AMPSO) algorithm was used to optimize the hyperparameters of the CNN model during training to obtain the AMPSO-CNN model. Finally, the partial convolutional layer parameters of the pre-trained network were transferred, and the remaining parameters were adjusted using real data samples to obtain the TL-AMPSO-CNN model. The trained model was used for gas-bearing distribution prediction, and its stability was assessed using uncertainty analysis. The results confirmed that the hybrid DL model performed efficiently in gas-bearing prediction compared to other individual models. Thus, for datasets with insufficient training samples, the proposed scheme can be applied as a feasible alternative to predict gas-bearing distribution.

A Novel Deep Learning-Based Data Analysis Model for Solar Photovoltaic Power Generation and Electrical Consumption Forecasting in the Smart Power Grid

With the installation of solar panels around the world and the permanent fluctuation of climatic factors, it is, therefore, important to provide the necessary energy in the electrical network in order to satisfy the electrical demand at all times for smart grid applications. This study first presents a comprehensive and comparative review of existing deep learning methods used for smart grid applications such as solar photovoltaic (PV) generation forecasting and power consumption forecasting. In this work, electrical consumption forecasting is long term and will consider smart meter data and socioeconomic and demographic data. Photovoltaic power generation forecasting is short term by considering climatic data such as solar irradiance, temperature, and humidity. Moreover, we have proposed a novel hybrid deep learning method based on multilayer perceptron (MLP), long short-term memory (LSTM), and genetic algorithm (GA). We then simulated all the deep learning methods on a climate and electricity consumption dataset for the city of Douala. Electrical consumption data are collected from smart meters installed at consumers in Douala. Climate data are collected at the climate management center in the city of Douala. The results obtained show the outperformance of the proposed optimized method based on deep learning in the both electrical consumption and PV power generation forecasting and its superiority compared to basic methods of deep learning such as support vector machine (SVM), MLP, recurrent neural network (RNN), and random forest algorithm (RFA).

A hybrid deep learning method for the prediction of ship time headway using automatic identification system data

Ship Time Headway (STH) is used in maritime navigation to describe the time interval between the arrivals of two consecutive ships in the same water area. This measurement may offer a straightforward way to gauge the frequency of ship traffic and the likelihood of congestion in a particular area. STH is an important factor in understanding and managing the dynamics of ship movements in busy waterways. This paper introduces a hybrid deep learning method for predicting STH in time domain. The method integrates the Seasonal-Trend Decomposition using Loess (STL), Multi-head Self-Attention (MSA) mechanism into Long Short-Term Memory (LSTM) neural network. The STH dataset was extracted from the Automatic Identification System (AIS) through ship trajectory spatial motion, and the seasonal, trend and residual components of the decomposition were then determined from the STH dataset using the STL algorithms. MSA-LSTM is adopted to comprehensively capture the evolving patterns of STH from the sequence. Comparison studies with existing methods demonstrate the accuracy and robustness of the predictions provided by this method, indicating that the proposed method outperforms other models in terms of prediction performance and learning capabilities. By predicting STH, the method offers potential to assist maritime traffic managers and navigators in assessing ship flow, thereby enabling them to make informed decisions on navigation safety and efficiency.

A hybrid deep learning approach for enhanced network intrusion detection

The contemporary era places paramount importance on network security and cloud environments, driven by increased data transmission demands, the flexibility of cloud services, and the prevalence of global resources. Addressing the escalating threat of computer malware, the development of efficient intrusion detection systems (IDS) is imperative. This research focuses on the challenges posed by imbalanced datasets and the necessity for unsupervised learning to enhance network security. The proposed hybrid deep learning method utilizes raw data from the CSE-CIC-IDS-2018 dataset, integrating imbalanced and unsupervised learning techniques. After preprocessing and normalization, feature extraction through principal component analysis (PCA) reduces dimensionality from seventy-eight fields to ten essential features. Clustering, employing the density-based spatial clustering of applications with noise (DBSCAN) algorithm optimized with particle swarm optimization (PSO), is applied to the extracted features, distinguishing between attack and non-attack packets. Addressing dataset imbalances, imbalanced learning techniques are employed, and unsupervised learning is exemplified through the AutoEncoder (AE) algorithm. The attack cluster’s data is input into AE, a deep learning-based approach, yielding outputs for attack classification. The proposed technique (PCA+DBSCANPSO+AE) achieves an impressive 99.19% accuracy in intrusion detection, surpassing contemporary methodologies and five existing techniques. This research not only enhances accuracy but also addresses imbalanced learning challenges, utilizing the power of unsupervised learning for robust network security.

Spatiotemporal Landsat-Sentinel-2 satellite imagery-based Hybrid Deep Neural network for paddy crop prediction using Google Earth engine

Rice is one of the predominant food sources to fulfill the dietary requirements of well-being in India. Therefore, accurate and timely paddy crop yield prediction is crucial to ensure the food security of the country. In this direction, the present study proposed a hybrid deep-learning method based on Conv-1D and LSTM layers using the classification-derived phenological with meteorological parameters for paddy crop yield prediction. The paddy crop classification has been conducted using high-resolution (10 m) multispectral imagery based on GPS coordinates collected during the paddy field visits to extract the phenological parameters for input to the prediction model. In this context, the efficiency of Random Forest, Naïve Bayes, SVM, and Gradient Tree boost classifiers was assessed. Furthermore, we have also analyzed the accuracy of Landsat-8, Sentinel-1 GRD, and Sentinel-2 satellite imagery in paddy crop classification based on area estimation. The Statistical Abstract of Haryana was utilized to validate the paddy crop area estimation and yield prediction. The classification outcomes showed that the Random Forest method attained the highest accuracy of 96.6 % compared to other GEE-based classifiers. The proposed Hybrid Deep learning approach achieved an RMSE value of 0.219 t/ha compared to CNN, LSTM, CNN-Bi-LSTM, and Regression techniques for crop yield prediction. The study conclusion highlighted that the sentinel-2 satellite imagery performed well and found that the proposed hybrid approach provided an alternative for paddy crop yield prediction.

An efficient approach based on a novel 1D-LBP for the detection of bearing failures with a hybrid deep learning method

Bearings serve as fundamental components in the transmission of motion for rotating machinery. The occurrence of mechanical wear and subsequent bearing failures within these rotating systems can lead to diminished operational efficiency and, if left unaddressed, may result in the complete cessation of the system's function. Hence, there exists a critical need for effective monitoring methodologies aimed at accurately detecting faults in such systems, preferably in their nascent stages. This study presents a novel approach to fault diagnosis leveraging vibration data obtained from bearings. Initially, a feature extraction technique is devised, which incorporates localized signal variations. Subsequently, these features, extracted via MM-1D-LBP, are utilized in conjunction with a hybrid deep learning network based on Long Short-Term Memory (LSTM) and one-dimensional Convolutional Neural Network (1D-CNN) architectures for diagnostic purposes. To assess the efficacy of the proposed methodology, experiments were conducted on two distinct datasets acquired from real-world bearing assemblies. In the first dataset, the aim was to predict various failure types (Inner Ring, Outer Ring, Ball). In the second dataset, the objective was to estimate defect sizes using bearing vibration signals corresponding to defects of different dimensions (0.15 cm, 0.5 cm, 0.9 cm) under consistent operating conditions. Remarkably high success rates of 99.31 % and 99.65 % were achieved for the two datasets, respectively, thus underscoring the efficacy of the proposed MM-1D-LBP+1D-CNN-LSTM approach. These findings not only demonstrate the feasibility of the proposed method for fault diagnosis in bearing systems but also suggest its potential applicability across diverse signal categories. Ultimately, this research contributes to advancing the state-of-the-art in fault diagnosis methodologies for rotating machinery, offering enhanced accuracy and early detection capabilities.

A hybrid LSTM method for forecasting demands of medical items in humanitarian operations

The inventory management of medical items in humanitarian operations is a challenging task due to the intermittent nature of their demand and long replenishment lead-times. An effective response to humanitarian emergencies is often associated with excess inventories, which leads to high costs. Henceforth, using accurate demand forecasts to control the inventories of these medical items is shown to be a significant lever to reduce inventory costs while keeping high service levels. This paper investigates the effectiveness of parametric and non-parametric demand forecasting methods that are commonly considered to deal with items characterised with intermittent demand patterns. Moreover, we propose a new hybrid deep learning method that combines the long short-term memory neural network and the support vector regression. Such combination enables to deal with complex and non-linear data. We conduct an empirical investigation by means of data related to 523 medical items managed in two warehouses of a major humanitarian organisation. The forecast accuracy and the inventory efficiency of the methods are analysed. The empirical results show the high performance of the deep learning methods. These insights are valuable not only in the context of medical items in humanitarian operations, but also for any items with intermittent demand patterns.

Increasing the Accuracy of Writer Identification Based on Bee Colony Optimization Algorithm and Hybrid Deep Learning Method

Increasing the Accuracy of Writer Identification Based on Bee Colony Optimization Algorithm and Hybrid Deep Learning Method

A hybrid deep learning method for distracted driving risk prediction based on spatio-temporal driving behavior data

ABSTRACT Timely exact distracted driving risk prediction is beneficial to perceive real-time traffic risk, which is an essential but challenging task in modern traffic safety management. The use and improvement of measures for road safety management will be better facilitated by grasping and analyzing the spatio-temporal patterns of driving behavior and forming predictions. In this paper, a Distracted Driving Risk Prediction (DDRP) neural network by deep learning and spatio-temporal dependence is proposed, which to accurately predict the scale of distracted driving behavior on road networks. Then, the method is employed for distracted driving risk prediction based on the provincial road network. The experiment demonstrates that our method performs relatively better than the other methods applied in this paper. In addition, the method can adapt to predict the scale of distracted driving behavior in different categories, time intervals, and grid cells.

A Hybrid Deep Learning Method for Short-Term Traffic Flow Forecasting: GSA-LSTM

A Hybrid Deep Learning Method for Short-Term Traffic Flow Forecasting: GSA-LSTM

A hybrid deep learning method for AE source localization for heterostructure of wind turbine blades

An acoustic emission (AE) and hybrid deep learning networks based damage source localization method for heterostructure of wind turbine blades is proposed in this paper. Firstly, comprehensive data preprocessing is performed, including AE signal denoising, feature extraction, feature selection and normalization. New training features including AE descriptors, features of time, frequency domains and spectral features are extracted. A feature selection method based on Light-GBM and correlation analysis is employed to identify relevant features for AE source localization. Subsequently, two deep learning networks, AM-BiLNN and AM-LCNN, are developed to locate the damage source in two steps. Then, numerical tests are implemented on localized structure of a wind turbine blade to verify the performance of the proposed method and the performance of the selected features and the robustness of the proposed method under noise are investigated. Furthermore, a comparative investigation between the proposed method with long short-term memory neural networks (LSTM), convolutional neural networks (CNN) and the cluster-based method is carried out to demonstrate the superiority of the proposed method. The results highlight the superiority and robustness of the proposed method. Feature selection is shown to effectively enhance coordinate localization performance.

Deterministic and probabilistic multi-time-scale forecasting of wind speed based on secondary decomposition, DFIGR and a hybrid deep learning method

Deterministic and probabilistic multi-time-scale forecasting of wind speed based on secondary decomposition, DFIGR and a hybrid deep learning method

Automated Spam Detection Using Sandpiper Optimization Algorithm-Based Feature Selection with the Machine Learning Model

The email has become an online communication tool and an important part of daily life. Spam mails take up a lot of space and bandwidth, and spam filtering algorithms have flaws that cause them to mistake legitimate emails for spam (false positives). These issues are becoming a bigger difficulty for the online world. This work proposes the use of a sandpiper optimization (SPO) algorithm, which is applied for the feature selection process which minimizes the training complexity and maximizes the classification accuracy and Radial Bias Neural Network (RBNN) for classifying emails as genuine email and spam email. The Enron email dataset and Spam Assassin datasets were used. The outcomes show that the rotation forest algorithm after feature selection with SPO accurately classifies the emails as genuine email and spam email with 3.88%, 5.75%, and 6.16% higher accuracy for Genuine Email, 2.31%, 8.47%, and 7.23% higher accuracy for spam email compared with existing Universal Spam Detection using Transfer Learning of the BERT Model (USD-TL-BERT), Hybrid Learning Approach for E-mail Spam Detection and Classification (HLA-ESDC), and the Email Spam Detection Using Hierarchical Attention Hybrid Deep Learning Method (ESD-HAHD), respectively. This demonstrates that the proposed method significantly outperforms existing methods.