Deep Neural Network Prediction Models Research Articles

A rapid prediction method for water droplet collection coefficients of multiple airfoils based on incremental learning and multi-modal dynamic fusion

The calculation of the water droplet collection coefficient (WDCC) is a crucial step in the numerical study of aircraft icing and the iterative design of anti-icing and deicing systems. Rapid and efficient methods for predicting WDCC are essential for enhancing the efficiency of icing numerical calculations and accelerating the design cycle of these systems. The existing prediction methods are inefficient and fail to meet the real-time requirements of engineering applications. This paper proposes a rapid prediction method for the WDCC for multiple airfoils utilizing deep learning techniques. The method takes enhanced airfoil section images and icing condition parameters as inputs and WDCC as output. A deep neural network prediction model, IncDynamicFusion, for sustainable learning is established by integrating a multimodal dynamic fusion method with an improved iCaRL method (incremental classifier and representation learning). Numerical experimental results demonstrate that the proposed model can quickly and effectively predict the WDCC of multiple airfoils.

ImageDTA: A Simple Model for Drug-Target Binding Affinity Prediction.

Predicting the drug-target binding affinity (DTA) is crucial in drug discovery, and an increasing number of researchers are using artificial intelligence techniques to make such predictions. Many effective deep neural network prediction models have been proposed. However, current methods need improvement in accuracy, complexity, and efficiency. In this study, we propose a method based on a multiscale 2-dimensional convolutional neural network (CNN), namely ImageDTA. Many studies have shown that CNN achieves good learning effects with limited data. Therefore, we take a unique perspective by treating the word vector encoded with a simplified molecular input line entry system (SMILES) string as an "image" and processing it like handling images, fully leveraging the efficient processing capabilities of CNN for image data. Furthermore, we show that ImageDTA has higher training and inference efficiency than pretrained large models and outperforms attention-based graph neural network models in accuracy and interpretability. We also use visualization techniques to select appropriate convolutional kernel sizes, thereby increasing the network's interpretability.

Multi-task aquatic toxicity prediction model based on multi-level features fusion

IntroductionWith the escalating menace of organic compounds in environmental pollution imperiling the survival of aquatic organisms, the investigation of organic compound toxicity across diverse aquatic species assumes paramount significance for environmental protection. Understanding how different species respond to these compounds helps assess the potential ecological impact of pollution on aquatic ecosystems as a whole. Compared with traditional experimental methods, deep learning methods have higher accuracy in predicting aquatic toxicity, faster data processing speed and better generalization ability. ObjectivesThis article presents ATFPGT-multi, an advanced multi-task deep neural network prediction model for organic toxicity. MethodsThe model integrates molecular fingerprints and molecule graphs to characterize molecules, enabling the simultaneous prediction of acute toxicity for the same organic compound across four distinct fish species. Furthermore, to validate the advantages of multi-task learning, we independently construct prediction models, named ATFPGT-single, for each fish species. We employ cross-validation in our experiments to assess the performance and generalization ability of ATFPGT-multi. ResultsThe experimental results indicate, first, that ATFPGT-multi outperforms ATFPGT-single on four fish datasets with AUC improvements of 9.8%, 4%, 4.8%, and 8.2%, respectively, demonstrating the superiority of multi-task learning over single-task learning. Furthermore, in comparison with previous algorithms, ATFPGT-multi outperforms comparative methods, emphasizing that our approach exhibits higher accuracy and reliability in predicting aquatic toxicity. Moreover, ATFPGT-multi utilizes attention scores to identify molecular fragments associated with fish toxicity in organic molecules, as demonstrated by two organic molecule examples in the main text, demonstrating the interpretability of ATFPGT-multi. ConclusionIn summary, ATFPGT-multi provides important support and reference for the further development of aquatic toxicity assessment. All of codes and datasets are freely available online at https://github.com/zhaoqi106/ATFPGT-multi.

SAMP-VGG16: Force-field assisted image-based deep neural network prediction model for short antimicrobial peptides.

During the last three decades, antimicrobial peptides (AMPs) have emerged as a promising therapeutic alternative to antibiotics. The approaches for designing AMPs span from experimental trial-and-error methods to synthetic hybrid peptide libraries. To overcome the exceedingly expensive and time-consuming process of designing effective AMPs, many computational and machine-learning tools for AMP prediction have been recently developed. In general, to encode the peptide sequences, featurization relies on approaches based on (a) amino acid (AA) composition, (b) physicochemical properties, (c) sequence similarity, and (d) structural properties. In this work, we present an image-based deep neural network model to predict AMPs, where we are using feature encoding based on Drude polarizable force-field atom types, which can capture the peptide properties more efficiently compared to conventional feature vectors. The proposed prediction model identifies short AMPs (≤30 AA) with promising accuracy and efficiency and can be used as a next-generation screening method for predicting new AMPs. The source code is publicly available at the Figshare server sAMP-VGG16.

A deep learning-based method for predicting the low-cycle fatigue life of austenitic stainless steel

In modern engineering, predicting the fatigue life of materials is crucial for safety assessment. The relationship between fatigue life and its influencing factors is difficult to predict by traditional methods, and deep learning can achieve great power and flexibility through nested hierarchies of concepts. Taking the low cycle fatigue life of 316 austenitic stainless steel as an example, a method for predicting the low cycle fatigue life of austenitic stainless steel by deep learning is established based on the limited ability of traditional neural network model and genetic algorithm optimization model. The deep neural network model is introduced to predict the fatigue life of the material. The results show that the prediction correlation coefficient R of the deep neural network prediction model with three hidden layers is 0.991, and the deep neural network learning model has better prediction ability.

Feature-based deep neural network approach for predicting mortality risk in patients with COVID-19

In this study, we integrate deep neural network (DNN) with hybrid approaches (feature selection and instance clustering) to build prediction models for predicting mortality risk in patients with COVID-19. Besides, we use cross-validation methods to evaluate the performance of these prediction models, including feature based DNN, cluster-based DNN, DNN, and neural network (multi-layer perceptron). The COVID-19 dataset with 12,020 instances and 10 cross-validation methods are used to evaluate the prediction models. The experimental results showed that the proposed feature based DNN model, holding Recall (98.62%), F1-score (91.99%), Accuracy (91.41%), and False Negative Rate (1.38%), outperforms than original prediction model (neural network) in the prediction performance. Furthermore, the proposed approach uses the Top 5 features to build a DNN prediction model with high prediction performance, exhibiting the well prediction as the model built by all features (57 features). The novelty of this study is that we integrate feature selection, instance clustering, and DNN techniques to improve prediction performance. Moreover, the proposed approach which is built with fewer features performs much better than the original prediction models in many metrics and can still remain high prediction performance.

Traditional Machine Learning Models for Building Energy Performance Prediction: A Comparative Research

A large proportion of total energy consumption is caused by buildings. Accurately predicting the heating and cooling demand of a building is crucial in the initial design phase in order to determine the most efficient solution from various designs. In this paper, in order to explore the effectiveness of basic machine learning algorithms to solve this problem, different machine learning models were used to estimate the heating and cooling loads of buildings, utilising data on the energy efficiency of buildings. Notably, this paper also discusses the performance of deep neural network prediction models and concludes that among traditional machine learning algorithms, GradientBoostingRegressor achieves better predictions, with Heating prediction reaching 0.998553 and Cooling prediction Compared with our machine learning algorithm HB-Regressor, the prediction accuracy of HB-Regressor is higher, reaching 0.998672 and 0.995153 respectively, but the fitting speed is not as fast as the GradientBoostingRegressor algorithm.

Ischemic stroke prediction of patients with carotid atherosclerotic stenosis via multi-modality fused network.

Carotid atherosclerotic stenosis of the carotid artery is an important cause of ischemic cerebrovascular disease. The aim of this study was to predict the presence or absence of clinical symptoms in unknown patients by studying the existence or lack of symptoms of patients with carotid atherosclerotic stenosis. First, a deep neural network prediction model based on brain MRI imaging data of patients with multiple modalities is constructed; it uses the multi-modality features extracted from the neural network as inputs and the incidence of diagnosis as output to train the model. Then, a machine learning-based classification algorithm is developed to utilize the clinical features for comparison and evaluation. The experimental results showed that the deep learning model using imaging data could better predict the clinical symptom classification of patients. As part of preventive medicine, this study could help patients with carotid atherosclerosis narrowing to prepare for stroke prevention based on the prediction results.

Facial Expression Guided Diagnosis of Parkinson's Disease via High-Quality Data Augmentation

Parkinson's disease (PD) is a neurodegenerative disease which is prevalent among the elder population and severely affects the life quality of patients and their families. Therefore, it is important to conduct an early diagnosis for potential patients with PD, so as to promote prompt treatment and avoid the aggravation of the disease. Recently, the in-vitro PD diagnosis based on facial expressions has received increasing attention because of its distinguishability (i.e., PD patients always possess the characteristics of “masked face”) and affordability. However, the performance of the existing facial expression-based PD diagnosis approaches is limited by: 1) the small-scale training data on PD patients' facial expressions, and 2) the weak prediction model. To address these two problems, we propose a new facial expression guided PD diagnosis method based on high-quality training data augmentation and deep neural network prediction. Specifically, the proposed method consists of three stages: Firstly, we synthesize virtual facial expression images with 6 basic emotions (i.e., anger, disgust, fear, happiness, sadness, and surprise) based on multi-domain adversarial learning to approximate the premorbid expressions of PD patients. Secondly, we introduce three facial image quality assessment (FIQA) criteria to measure the quality of these synthesized facial expression images and design a fusion screening strategy that shortlists the high-quality ones to augment the training data. Finally, we train a deep neural network prediction model based on the original and synthesized high-quality facial expression images for PD diagnosis. To show real-world impacts and evaluate the proposed method under different facial expressions, we also create a (currently largest) multiple facial expressions-based PD face dataset in collaboration with a hospital. Extensive experiments are performed to demonstrate the effectiveness of the multi-domain adversarial learning-based facial expression synthesis and the fusion screening strategy, particularly the superior performance of the proposed method for PD diagnosis.

Machine learning algorithm and deep neural networks identified a novel subtype in hepatocellular carcinoma.

Hepatocellular carcinoma (HCC) is one of the most common malignant tumors. Due to the lack of specific characteristics in the early stage of the disease, patients are usually diagnosed in the advanced stage of disease progression. This study used machine learning algorithms to identify key genes in the progression of hepatocellular carcinoma and constructed a prediction model to predict the survival risk of HCC patients. The transcriptome data and clinical information were downloaded from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO). The differential expression analysis and COX proportional-hazards model participated in the identification of survival-related genes. K-Means, Random forests, and LASSO regression are involved in identifying novel subtypes of HCC and screening key genes. The prediction model was constructed by deep neural networks (DNN), and Gene Set Enrichment Analysis (GSEA) reveals the metabolic pathways where key genes are located. Two subtypes were identified with significantly different survival rates (p< 0.0001, AUC = 0.720) and 17 key genes associated with the subtypes. The accuracy rate of the deep neural network prediction model is greater than 93.3%. The GSEA analysis found that the survival-related genes were significantly enriched in hallmark gene sets in the MSigDB database. In this study, we used machine learning algorithms to screen out 17 genes related to the survival risk of HCC patients, and trained a DNN model based on them to predict the survival risk of HCC patients. The genes that make up the model are all key genes that affect the formation and development of cancer.

Deep Neural Network Prediction Model of Hydrogen Content in VOD Process Based on Small Sample Dataset

Deep Neural Network Prediction Model of Hydrogen Content in VOD Process Based on Small Sample Dataset

The Prediction of Enterprise Stock Change Trend by Deep Neural Network Model.

This study aims to accurately predict the changing trend of stocks in stock trading so that company investors can obtain higher returns. In building a financial forecasting model, historical data and learned parameters are used to predict future stock prices. Firstly, the relevant theories of stock forecasting are discussed, and problems in stock forecasting are raised. Secondly, the inadequacies of deep neural network (DNN) models are discussed. A prediction trend model of enterprise stock is established based on long short-term memory (LSTM). The uniqueness and innovation lie in using the stock returns of Bank of China securities in 2022 as the training data set. LSTM prediction models are used to perform error analysis on company data training. The 20-day change trend of the company's stock returns under different models is predicted and analyzed. The results show that as the number of iterations increases, the loss rate of the LSTM training curve keeps decreasing until 0. The average return price of the LSTM prediction model is 14.01. This figure is closest to the average real return price of 13.89. Through the forecast trend analysis under different models, LSTM predicts that the stock change trend of the enterprise model is closest to the changing trend of the actual earnings price. The prediction accuracy is better than other prediction models. In addition, this study explores the characteristics of high noise and complexity of corporate stock time series, designs a DNN prediction model, and verifies the feasibility of the LSTM model to predict corporate stock changes with high accuracy.

Receding horizon control based energy management strategy for PHEB using GRU deep learning predictive model

As an important method to achieve clean public transportation, plug-in hybrid electric bus (PHEB) has been gradually utilized in the transport system. The energy consumption performance of PHEB is mainly determined by its energy management strategy (EMS). Aiming at improving fuel economy of PHEB and making better use of battery, a receding horizon control (RHC) based EMS is proposed in this paper. Under the real driving cycles, the gated recurrent units (GRU) based model is utilized to predict the velocity sequence over receding horizon, and two other deep neural network prediction models are introduced for comparison. Moreover, the trip distance-based state of charge (SOC) constraint method is designed. Combining the velocity predictor and the SOC constraints, the power distribution of PHEB is described as a rolling optimization problem in the prediction horizon. The simulation and hardware-in-loop (HIL) experiments demonstrate that this strategy can improve fuel economy while ensuring the rational battery discharging under different initial SOC and traveling distance.

A Hybrid Stock Price Prediction Model Based on PRE and Deep Neural Network

Stock prices are volatile due to different factors that are involved in the stock market, such as geopolitical tension, company earnings, and commodity prices, affecting stock price. Sometimes stock prices react to domestic uncertainty such as reserve bank policy, government policy, inflation, and global market uncertainty. The volatility estimation of stock is one of the challenging tasks for traders. Accurate prediction of stock price helps investors to reduce the risk in portfolio or investment. Stock prices are nonlinear. To deal with nonlinearity in data, we propose a hybrid stock prediction model using the prediction rule ensembles (PRE) technique and deep neural network (DNN). First, stock technical indicators are considered to identify the uptrend in stock prices. We considered moving average technical indicators: moving average 20 days, moving average 50 days, and moving average 200 days. Second, using the PRE technique-computed different rules for stock prediction, we selected the rules with the lowest root mean square error (RMSE) score. Third, the three-layer DNN is considered for stock prediction. We have fine-tuned the hyperparameters of DNN, such as the number of layers, learning rate, neurons, and number of epochs in the model. Fourth, the average results of the PRE and DNN prediction model are combined. The hybrid stock prediction model results are computed using the mean absolute error (MAE) and RMSE metric. The performance of the hybrid stock prediction model is better than the single prediction model, namely DNN and ANN, with a 5% to 7% improvement in RMSE score. The Indian stock price data are considered for the work.

Retracted] Research on the Prediction of Port Economic Synergy Development Trend Based on Deep Neural Networks

After entering the new century, along with the further deepening of global economic and trade cooperation, the industrial division of labor has been newly developed globally, which brings the cooperation among countries in the international logistics chain more and more closely. As the core node linking domestic and foreign water transportation, ports play a very key role in the international logistics chain and have an extremely central position in the national logistics planning. The coordinated development of port economy is an important part of the economic development planning of port cities, and it is also the premise and basis for the comprehensive planning of port logistics infrastructure construction scale, logistics space layout, and port city logistics development direction and function positioning. Therefore, according to the availability of realistic data, this paper establishes a deep neural network prediction model for the collaborative development of ports and uses various port logistics indicators to predict the economic development trend, so as to realize a nonlinear mapping relationship between the level of port economic development and the side of port logistics demand. Meanwhile, the research of this paper will provide theoretical basis and corresponding practical tools for the coordinated development between regional economy and port logistics industry.

A Novel Hybrid Deep Neural Network Prediction Model for Shield Tunneling Machine Thrust

Shield thrust is a critical operational parameter during shield driving, which is of vital significance for adjusting operational parameters and ensuring efficient and safe propulsion of shield tunneling machine. In this paper, a novel hybrid prediction model (CLM) combining attention mechanism, convolutional neural networks (CNN) and Bi-directional long short-term memory (BiLSTM) network is proposed for shield thrust prediction. Correlation analysis based on Maximal Information Coefficient (MIC) between the thrust and other parameters is first conducted to select optimal parameters and reduce input dimension. An attention mechanism is introduced into CNN to distinguish the importance of different features, with the convolution layer and pooling layer further extracting dimension features of the data. Then, a BiLSTM neural network integrating first attention layer is employed to extract time-varying characteristics of the data, with a second attention layer added to capture important time information. Field data during shield cutting bridge piles are investigated to support and validate the effectiveness and superiority of the proposed method. Results show that the proposed CLM model are general enough to avoid overfitting problems and have good performance at prediction. The predicted value match reasonably well the monitoring data, with coefficient of determination (R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> ) equaling to 0.85, root mean square error (RMSE) equaling to 0.05, mean absolute error (MAE) equaling to 0.02. The CLM model in this paper can accurately predict the thrust even under complicated construction conditions, which provides reference for similar industrial application.

Multi-objective aerodynamic optimization design of variable camber leading and trailing edge of airfoil

Variable camber is an effective method for improving the flight efficiency of large aircraft, and has attracted the attention of researchers. This work focused on the optimization of a variable camber airfoil. First, the influences of the variable camber of the leading and trailing edges on the airfoil aerodynamic performance were investigated using a computational fluid dynamics numerical simulation. An initial database was established for a deep neural network. Second, an iterative algorithm was constructed to optimize the variable camber airfoil in terms of the rotation angle of the leading edge, deflection position of the leading edge, rotation angle of the trailing edge, and deflection position of the trailing edge. A genetic algorithm was used in each iteration to maximize the lift coefficient and lift-to-drag ratio, as predicted using a deep neural network (DNN). The optimal results were validated using Fluent. If the DNN result approximated the Fluent results, the iterative process was stopped. Otherwise, the Fluent results were inserted into the database to update the DNN prediction model. The optimization results showed that the lift-to-drag ratio of the 2D airfoil could be increased by more than 14 when the angle of attack was less than 8° relative to the original airfoil. Furthermore, to validate the 2D optimal results, the optimized 2D airfoil was stretched into 3D, and it was discovered that the aerodynamic performance trend of the 3D airfoil with respect to the angle of attack was basically the same as that of the 2D airfoil. In addition, the corresponding 3D airfoil improved the aerodynamic performance and reduced the noise at a high frequency (by approximately 16 dB). In contrast, the noise in the low and medium frequencies remained unchanged. Therefore, the optimization method and results can provide a reference for the aerodynamic design and acoustic design of large civil aircraft wings.

DEM study of flow characteristics of wet cohesive particles in packed bed

The study of particle discharge flow is significant in industrial processes,especially wet cohesive particles. In this paper, the discrete element method (DEM), cohesion model and the viscous force model are used to study the influencing parameters of the discharge flow characteristics of cohesive system in three-dimensional packed bed. The cohesion between particles is described by liquid bridge model, the Bond number Bo and liquid content W are used to study the cohesive behaviour. The discharge efficiency decreases with the increase of particle viscosity. When Bo⩽0.1 or W⩽6%, the uniformity of the particles in the hopper along the direction of gravity is basically the same of non-cohesive particle, but cohesive particles have a greater horizontal velocity fluctuations in the central area. The residence time tr of particles increases with the increase of cohesive force. The research results help to better understand the discharge characteristics of cohesive particle and provide a reference for the discharge behavior of the reactor core, especially under adverse damage conditions. Based on the simulation data, a Deep Neural Network (DNN) prediction model of residence time is proposed. The DNN prediction model was proposed to train the simulated data and predict the residence time of the particles in hopper based on the initial coordinates of each particle. After training and verification on more than 600,000 data, it has reached an accuracy rate of over 98% and meets the needs of engineering calculation.

The adoption of deep neural network (DNN) to the prediction of soil liquefaction based on shear wave velocity

Soil liquefaction has been accepted as one of the factors causing natural disasters and engineering failures in the seismic. The mathematic prediction model for soil liquefaction is widely accepted, and the standard penetration (SPT) and cone penetration test (CPT) prediction model using the machine learning method is also developed. But for the Vs, the prediction model based on the machine learning method is limited. So, considering the advantage of the deep learning method, a multi-layer fully connected network (ML-FCN) was proposed to optimize the deep neural network (DNN) and adopted to train the prediction model based on the Vs and SPT dataset in this paper. The history dataset was divided into a training set, a validation set, and a testing set by a ratio of 6:2:2 for better evaluation. The SPT dataset was extracted to train a corresponding DNN prediction model. According to the comparison results, the model trained by ML-FCN DNN could predict the liquefaction potential with higher accuracy than the model proposed by Hanna et al. (Soil Dyn Earthq Eng 27(6):521–40, 2007), which is enough to be applied to real engineering, the parameter of Vs is essential to improve the model performance as for the three sets.

Predictors of 30-Day Unplanned Readmission After Carotid Artery Stenting Using Artificial Intelligence.

IntroductionThis study aimed to describe the rates and causes of unplanned readmissions within 30 days following carotid artery stenting (CAS) and to use artificial intelligence machine learning analysis for creating a prediction model for short-term readmissions. The prediction of unplanned readmissions after index CAS remains challenging. There is a need to leverage deep machine learning algorithms in order to develop robust prediction tools for early readmissions.MethodsPatients undergoing inpatient CAS during the year 2017 in the US Nationwide Readmission Database (NRD) were evaluated for the rates, predictors, and costs of unplanned 30-day readmission. Logistic regression, support vector machine (SVM), deep neural network (DNN), random forest, and decision tree models were evaluated to generate a robust prediction model.ResultsWe identified 16,745 patients who underwent CAS, of whom 7.4% were readmitted within 30 days. Depression [p < 0.001, OR 1.461 (95% CI 1.231–1.735)], heart failure [p < 0.001, OR 1.619 (95% CI 1.363–1.922)], cancer [p < 0.001, OR 1.631 (95% CI 1.286–2.068)], in-hospital bleeding [p = 0.039, OR 1.641 (95% CI 1.026–2.626)], and coagulation disorders [p = 0.007, OR 1.412 (95% CI 1.100–1.813)] were the strongest predictors of readmission. The artificial intelligence machine learning DNN prediction model has a C-statistic value of 0.79 (validation 0.73) in predicting the patients who might have all-cause unplanned readmission within 30 days of the index CAS discharge.ConclusionsMachine learning derived models may effectively identify high-risk patients for intervention strategies that may reduce unplanned readmissions post carotid artery stenting.Central IllustrationFigure 2: ROC and AUPRC analysis of DNN prediction model with other classification models on 30-day readmission data for CAS subjectsSupplementary InformationThe online version contains supplementary material available at 10.1007/s12325-021-01709-7.