Long Short-term Memory Network Research Articles

Cucumber Downy Mildew Disease Prediction Using a CNN-LSTM Approach

It is of great significance to develop early prediction technology for controlling downy mildew and promoting cucumber production. In this study, a cucumber downy mildew prediction method was proposed by fusing quantitative disease information and environmental data. Firstly, the number of cucumber downy mildew spores during the experiment was collected by a portable spore catcher, and the proportion of cucumber downy mildew leaf area to all cucumber leaf area was recorded, which was used as the incidence degree of cucumber plants. The environmental data in the greenhouse were monitored and recorded by the weather station in the greenhouse. Environmental data outside the greenhouse were monitored and recorded by a weather station in front of the greenhouse. Then, the influencing factors of cucumber downy mildew were analyzed based on the Pearson correlation coefficient method. The influencing factors of the cucumber downy mildew early warning model in greenhouse were identified. Finally, the CNN-LSTM (Convolutional Neural Network-Long Short-Term Memory) algorithm was used to establish the cucumber downy mildew incidence prediction model. The results showed that the Mean Absolute Error (MAE), Mean Square Error (MSE), Root Mean Square Error (RMSE), and determination coefficient (R2) of the CNN-LSTM network model were 0.069, 0.0098, 0.0991, and 0.9127, respectively. The maximum error between the predicted value and the true value for all test sets was 16.9398%. The minimum error between the predicted value and the true value for all test sets was 0.3413%. The average error between the predicted and true values for all test sets was 6.6478%. The Bland–Altman method was used to analyze the predicted and true values of the test set, and 95.65% of the test set data numbers were within the 95% consistency interval. This work can serve as a foundation for the creation of early prediction models of greenhouse crop airborne diseases.

Study on updating finite element model of steel truss structure based on knowledge-enhanced deep reinforcement learning

Accurate assessment of the damage to a structure is the key to structural health monitoring. Among them, one effective means to assess the real-time status of the structure is by establishing a finite element model (FEM) of the structure and updating it synchronously. Previous research mainly updates the FEM of the structure with a certain moment in time without considering the continuous changes of structural parameters over a longer time. In this paper, we propose a FEM updating method for steel truss structure based on knowledge-enhanced deep reinforcement learning, which introduces long short-term memory (LSTM) network based on a deep deterministic policy gradient (DDPG) algorithm and constructs a knowledge-enhanced noise function (NF) module based on parametric analysis. The NF-LSTM-DDPG algorithm can fit a continuous variation of structural parameters. The ablation experiments revealed that the LSTM network and the NF module facilitate the exploration and learning ability of the DDPG algorithm. The NF-LSTM-DDPG algorithm is used in an actual steel truss structure case, and the variation rules of the elastic modulus and support deviations obtained from the solution are consistent with the actual situation. The average relative error of the 34 strain measurement points after updating is 14.147 %, and the relative error is higher than 20 % in only 13 days out of 187 days of continuous monitoring. Five days were selected for the comparative study, and the experimental results show that the relative error of the NF-LSTM-DDPG algorithm is 13.206 %, which is higher than that of the particle swarm optimization algorithm and the simulated annealing algorithm. The change of elastic modulus updated based on the NF-LSTM-DDPG algorithm is more reasonable than other algorithms.

Forecasting sea surface temperature during typhoon events in the Bohai Sea using spatiotemporal neural networks

Advanced neural network methods can effectively combine temporal and spatial information, allowing us to exploit complex relationships in the prediction of sea surface temperature (SST). Therefore, the authors have developed an attention-based context fusion network model recently, known as ACFN, to enhance the underlying spatiotemporal correlations in SST data. In a previous preliminary long-term evaluation, ACFN has demonstrated substantial improvements over several state-of-the-art baseline models, such as Convolutional Long-Short Term Memory (ConvLSTM) and Predictive Recurrent Neural Network (PredRNN). However, these methods have not been widely used for predicting SST during typhoons, and their performance has not been thoroughly evaluated. This study focuses on three specific typhoons, In-fa (2021), Lekima (2019), and Rumbia (2018), as case studies. The results show that ACFN consistently outperforms ConvLSTM and PredRNN, as demonstrated during Typhoon In-fa with a mean absolute error of 0.27 °C and a root-mean-square error of 0.33 °C for a 1-day forecast time. This study also examines the impact of forecast time on predictive skill, revealing that the performance of ACFN, as expected, gradually decreases with longer forecast times. This thorough evaluation of the ACFN model provides a valuable reference for using deep learning in predicting regional typhoon SST. Plain language summaryIn previous work, we developed an attention-based context fusion network model. This spatiotemporal neural network model has been successfully applied to short-term sea surface temperature (SST) prediction in the Bohai Sea, utilizing only current and previous 1- to 10-day SST data as input to forecast SST for the subsequent 1- to 10-day period. In this study, we further evaluated the capability of the model to predict SST during typhoons in the Bohai Sea, and the results were very promising. When predicting SST during Typhoon In-fa, the present spatiotemporal neural network model had a mean absolute error of only 0.27 °C for a 1-day forecast time, and a correlation coefficient of 0.74. On the other hand, comparison with baseline spatiotemporal neural network models shows that the present model is more effective in extracting input information. The current model has resolved the constraint that its predictive capability does not diminish as the forecast time increases, thereby enhancing the rationality of the prediction outcomes. In conclusion, the present spatiotemporal neural network model demonstrates its efficacy in forecasting typhoon-induced SST in the Bohai Sea, offering valuable insights for applications in meteorology and oceanography.

Prediction of Daily Climate Using Long Short-Term Memory (LSTM) Model

Climaate prediction plays a vital role in various sectors, including agriculture, disaster management, and urban planning. Traditional methods for climate forecasting often rely on complex physical models, which require substantial computational resources and may not accurately capture local weather patterns. This study explores the potential of Long Short-Term Memory (LSTM) networks, a type of recurrent neural network, for predicting daily climate variables such as temperature, precipitation, and humidity. Utilizing historical climate data from the city of Delhi, we developed an LSTM model to forecast short-term climate trends. The model consists of two LSTM layers followed by three Dense layers and is compiled with the Adam optimizer, mean squared error loss, and mean absolute error as a metric. Our results demonstrate the model's capability to capture temporal dependencies in climate data, achieving a satisfactory level of accuracy in temperature forecasting. This research underscores the potential of machine learning techniques, particularly LSTM networks, in enhancing climate prediction and contributing to more informed decision-making in weather-sensitive sectors.

Using Crested Porcupine Optimizer Algorithm and CNN-LSTM-Attention Model Combined with Deep Learning Methods to Enhance Short-Term Power Forecasting in PV Generation

Due to the inherent intermittency, variability, and randomness, photovoltaic (PV) power generation faces significant challenges in energy grid integration. To address these challenges, current research mainly focuses on developing more efficient energy management systems and prediction technologies. Through optimizing scheduling and integration in PV power generation, the stability and reliability of the power grid can be further improved. In this study, a new prediction model is introduced that combines the strengths of convolutional neural networks (CNNs), long short-term memory (LSTM) networks, and attention mechanisms, so we call this algorithm CNN-LSTM-Attention (CLA). In addition, the Crested Porcupine Optimizer (CPO) algorithm is utilized to solve the short-term prediction problem in photovoltaic power generation. This model is abbreviated as CPO-CLA. This is the first time that the CPO algorithm has been introduced into the LSTM algorithm for parameter optimization. To effectively capture univariate and multivariate time series patterns, multiple relevant and target variables prediction patterns (MRTPPs) are employed in the CPO-CLA model. The results show that the CPO-CLA model is superior to traditional methods and recent popular models in terms of prediction accuracy and stability, especially in the 13 h timestep. The integration of attention mechanisms enables the model to adaptively focus on the most relevant historical data for future power prediction. The CPO algorithm further optimizes the LSTM network parameters, which ensures the robust generalization ability of the model. The research results are of great significance for energy generation scheduling and establishing trust in the energy market. Ultimately, it will help integrate renewable energy into the grid more reliably and efficiently.

Evaluation of fish meal freshness using a metal-oxide semiconductor electronic nose combined with the long short-term memory feature extraction method.

To improve the classification and regression performance of the total volatile basic nitrogen (TVB-N) and acid value (AV) of different freshness fish meal samples detected by a metal-oxide semiconductor electronic nose (MOS e-nose), 402 original features, 62 manually extracted features, manually extracted and selected features by the RFRFE method, and the features extracted by the long short-term memory (LSTM) network were used as inputs to identify the freshness. The classification performance of the freshness grades and the estimation performance of the TVB-N and AV values of fish meal with different freshness were compared. According to the sensor response curve, preprocessing and feature extraction steps were first applied to the original data. Then, five classification algorithms and four regression algorithms were used for modeling. The results showed that a total of 30 features were extracted using the LSTM network, and the number of extracted features was significantly reduced. In the classification, the highest accuracy rate of 95.4% was obtained using the support vector machine method. In the regression, the least squares support vector regression method obtained the best root mean square error (RMSE). The coefficient of determination (R2), RMSE, and relative standard deviation (RSD) between the predicted value of TVBN and the actual value were 0.963, 11.01, and 7.9%, respectively. The R2, RMSE, and RSD between the predicted value of AV and the actual value were 0.972, 0.170, and 6.05%, respectively. The LSTM feature extraction method provided a new method and reference for feature extraction using an E-nose to identify other animal-derived material samples.

Survey of short-term traffic flow prediction based on LSTM

Short-time traffic flow prediction can not only help urban traffic management to complete the control and induce but also reduce the degree of urban congestion and improve the efficiency of urban operation. At the same time, improving the effect of short-time traffic flow prediction is one of the key points of traffic control and guidance development. Currently, there are many different models for short-term traffic flow prediction, and different prediction models have their different advantages and disadvantages. In this paper, we review the current status of research on short-time traffic flow prediction based on Long Short-Term Memory (LSTM) neural network: first, we analyze the research method in this paper; second, we briefly review the research and application effects of common prediction methods in traffic flow prediction; finally, the optimization method based on LSTM and its application effect are summarized and analyzed from the aspect of LSTM network and its combination model. In addition, short-time traffic flow prediction still faces challenges due to the problems of the data itself and the complexity of real-world traffic flow impact factors.

An EEMD‐LSTM, SVR, and BP decomposition ensemble model for steel future prices forecasting

AbstractThe forecasting of steel futures prices is important for the steel futures market, even for the steel industry. We propose a decomposition ensemble model that incorporates the Ensemble Empirical Mode Decomposition (EEMD), Long Short‐Term Memory (LSTM), Support Vector Regression (SVR), and Back Propagation (BP) neural network to forecast steel futures prices. The forecasting procedures are as follows: (1) The price data are initially decomposed into several relatively independent Intrinsic Mode Functions (IMFs) and a residue using EEMD. (2) The IMFs are then reconstructed as components representing short‐term, medium‐term, and long‐term frequencies via fine‐to‐coarse. (3) LSTM, SVR, and BP neural network are utilized to forecast the short‐term, medium‐term, and long‐term reconstructed components, respectively. (4) The prediction results for each component are simply added to the final prediction results. The accuracy of the proposed model is compared with several benchmark models by experiments and evaluated by some prediction evaluation indexes. The experimental results show that our model outperforms other models in terms of forecast accuracy, confirming its strong predictive capabilities. This study provides some suggestions for investment and decision making by participants in the steel futures market. It may promote the smooth operation of the steel futures market and shed some light on the operation of the steel industry.

Physics-Informed machine learning for tribological properties prediction of S32750/CFRPEEK tribopair under seawater lubrication via PISSA-CNN-LSTM

Tribological properties and wear mechanisms of S32750 stainless steel/carbon fiber-reinforced polyetheretherketone (CFRPEEK) tribopair in seawater under different salinities, sliding speeds and loads were experimentally studied. A novel Physics-Informed constraint strategy has been originally developed and combined with sparrow search algorithm (SSA) to optimize a Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) model for predicting friction coefficient. The results indicate that an increase in seawater salinity and sliding speed enhances the tribological performance. Under high load, an increase in the PEEK transfer film leads to a slight decrease in friction coefficient. The Physics-Informed SSA (PISSA) offers enhanced performance through faster convergence, higher solution quality, and less computation time. Additionally, CNN-LSTM model optimized by PISSA (PISSA-CNN-LSTM) provides more accurate predictions of friction coefficient.

Early Detection of Potato Disease Using an Enhanced Convolutional Neural Network-Long Short-Term Memory Deep Learning Model

Early Detection of Potato Disease Using an Enhanced Convolutional Neural Network-Long Short-Term Memory Deep Learning Model

Multi-Shared-Task Self-Supervised CNN-LSTM for Monitoring Free-Body Movement UPDRS-III Using Wearable Sensors.

The Unified Parkinson's Disease Rating Scale (UPDRS) is used to recognize patients with Parkinson's disease (PD) and rate its severity. The rating is crucial for disease progression monitoring and treatment adjustment. This study aims to advance the capabilities of PD management by developing an innovative framework that integrates deep learning with wearable sensor technology to enhance the precision of UPDRS assessments. We introduce a series of deep learning models to estimate UPDRS Part III scores, utilizing motion data from wearable sensors. Our approach leverages a novel Multi-shared-task Self-supervised Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) framework that processes raw gyroscope signals and their spectrogram representations. This technique aims to refine the estimation accuracy of PD severity during naturalistic human activities. Utilizing 526 min of data from 24 PD patients engaged in everyday activities, our methodology demonstrates a strong correlation of 0.89 between estimated and clinically assessed UPDRS-III scores. This model outperforms the benchmark set by single and multichannel CNN, LSTM, and CNN-LSTM models and establishes a new standard in UPDRS-III score estimation for free-body movements compared to recent state-of-the-art methods. These results signify a substantial step forward in bioengineering applications for PD monitoring, providing a robust framework for reliable and continuous assessment of PD symptoms in daily living settings.

Multidimensional Forensic Analysis of Onion Sites Using LSTM and Fuzzy Encoded Techniques: Insights and Applications in Digital Investigations

When it comes to digital investigations, the fact that onion sites make people anonymous makes investigative research much harder. For multidimensional forensic analysis of onion sites, this study suggests a new method that combines Long Short-Term Memory (LSTM) networks with fuzzy encoding methods. When you use the Tor network to get to onion sites, both users and services become anonymous, which makes it harder to do regular investigations. LSTM networks are used to look at trends in how people use the site and when material is updated over time. These networks are good at modeling sequential data. Fuzzy encoding methods work with LSTM to handle the uncertainty that comes with analyzing anonymous data, which makes investigative results more accurate. The method involves keeping time logs of what people do on onion sites and turning them into fuzzy sets to show how user interactions and content changes change over time. LSTM models are taught on these recorded sequences to find patterns that point to suspicious activity, like sudden changes in content or strange patterns of user access. This method can be used to find illegal activities, find secret services that host illegal content, and make profiles of how users behave in an anonymous network. Investigators can better track actions across onion sites when they use LSTM's ability to see how events depend on each other over time and fuzzy encoding's ability to deal with inaccurate data. Real-world onion site data used for experiments shows that the proposed method can find secret behaviors and patterns that are hard for traditional investigative methods to find. The results show that the accuracy and efficiency of detecting are much better than with traditional methods.

A data-driven LSTMSCBLS model for soft sensor of industrial process

Abstract In the chemical industry, data-driven soft sensor modeling plays a crucial role in efficiently monitoring product quality and status. Industrial data in these applications typically exhibit significant temporal characteristics, meaning that current information is influenced by data from previous periods. Effectively extracting and utilizing these temporal features is essential for achieving accurate soft sensor modeling in complex chemical scenarios. To address this challenge, this study proposes a data-driven Broad Learning System (BLS) model, which combines Long Short-Term Memory (LSTM) networks with an adaptive algorithm known as the Stochastic Configuration Algorithm (SC), referred to as LSTMSCBLS. The model operates in two stages: temporal feature extraction and final prediction. In the temporal feature extraction stage, the integration of the LSTM network with a feature attention mechanism allows for efficient extraction of temporal features from high-dimensional time-series data. In the final prediction stage, the SC is integrated into the BLS, effectively mitigating issues related to node space redundancy and the determination of the number of nodes. The effectiveness and superiority of the proposed model are demonstrated through two industrial case studies involving a debutanizer column and a sulfur recovery unit.

Data-efficient comfort modeling: Active transfer learning for predicting personal thermal comfort using limited data

Personal comfort models are used to predict thermal comfort responses at the individual level rather than predicting the average thermal comfort responses for large populations. These models, data-driven in nature, need to be trained on large amounts of occupant comfort feedback and sensor data to achieve accurate predictions. However, collecting such data is often expensive and labor-intensive in reality. To address this, we proposed a data-efficient active transfer learning (ATL) framework to improve the performance of personal comfort models under limited data. To demonstrate the validity of this framework, we developed a base Convolutional Neural Network-Long Short-term Memory (CNNLSTM) model alongside two transfer learning models utilizing feature extraction (TL-CNNLSTM-FE) and fine-tuning (TL-CNNLSTM-FE) approaches, enhanced by a novel active learning strategy. Using these models, three comfort prediction tasks (i.e., thermal preference, thermal acceptability, and air movement preference) were performed by transferring the knowledge from the ASHRAE Global Thermal Comfort Database II to a limited dataset collected in the tropics. Empirical results indicate that the active transfer learning framework proposed was able to consistently outperform the base and transfer learning models using only less than 10% of the training data for all personal comfort tasks, highlighting the effectiveness of this strategy. The implications of this work are especially useful for the research community working on the practical applications of data-efficient machine learning approaches for personal thermal comfort predictions.

Adaptation of an alternative neural network instead of ANN for backcalculating pavement layer moduli

The structural condition of the pavement is often evaluated using deflection data from the non-destructive test such as the Falling Weight Deflectometer (FWD). Young's moduli calculated from surface deflections are used to characterize pavement layers. One of the most popular techniques for analyzing FWD data to determine pavement layer moduli is backcalculation. Artificial Neural Network (ANN) has been explored or used in studies to backcalculate layer moduli from FWD data, although its applications in this area are relatively recent. However, ANN suffers from limitations in terms of convergence accuracy and generalization capability. This research aims to develop an alternative neural network instead of ANN for backcalculating pavement layer moduli. Based on FWD data, Recurrent Neural Network (RNN) and Long-Short Term Memory (LSTM) networks were used to overcome the shortcomings of the traditional ANN. The LSTM, RNN, and ANN networks were developed and trained with the same training properties. The results revealed that LSTM attained high convergence accuracy and fastest convergence speed than ANN and RNN. LSTM network produced reasonable and accurate layer moduli values with determination coefficient (R2) of 0.9298 when compared to the measured values, while R2 values of the RNN and ANN networks were 0.906 and 0.8326, respectively. There was evidence that the LSTM network can learn the continuity pattern between the deflection basin points and enhance the FWD backcalculation accuracy.

Exploring the Relationship and Predictive Accuracy for the Tadawul All Share Index, Oil Prices, and Bitcoin Using Copulas and Machine Learning

Financial markets are increasingly interlinked. Therefore, this study explores the complex relationships between the Tadawul All Share Index (TASI), West Texas Intermediate (WTI) crude oil prices, and Bitcoin (BTC) returns, which are pivotal to informed investment and risk-management decisions. Using copula-based models, this study identified Student’s t copula as the most appropriate one for encapsulating the dependencies between TASI and BTC and between TASI and WTI prices, highlighting significant tail dependencies. For the BTC–WTI relationship, the Frank copula was found to have the best fit, indicating nonlinear correlation without tail dependence. The predictive power of the identified copulas were compared to that of Long Short-Term Memory (LSTM) networks. The LSTM models demonstrated markedly lower Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Scaled Error (MASE) across all assets, indicating higher predictive accuracy. The empirical findings of this research provide valuable insights for financial market participants and contribute to the literature on asset relationship modeling. By revealing the most effective copulas for different asset pairs and establishing the robust forecasting capabilities of LSTM networks, this paper sets the stage for future investigations of the predictive modeling of financial time-series data. The study highlights the potential of integrating machine-learning techniques with traditional econometric models to improve investment strategies and risk-management practices.

Butterfly valve erosion prediction based on LSTM network

When valves and pipelines are used, erosion wear is a major concern. Erosion wear can lead to equipment downtime, material replacement, and other issues, as well as seal surface failure. The research delves into the phenomenon of erosion in butterfly valves, crucial components utilized across diverse industrial sectors. Erosion primarily affects the valve's disc and seat regions, leading to premature wear and compromised performance. To address this issue, the research employs computational fluid dynamics (CFD) and machine learning techniques, including long short-term memory (LSTM) and BP neural networks, to predict erosion rates under various conditions (different valve openings and particle diameters). Results indicate that erosion escalates with time and larger valve openings, highlighting the need for proactive maintenance strategies. The LSTM model demonstrates superior predictive capabilities compared to the BP neural network, offering valuable insights for improving butterfly valve design and operational efficiency in industrial settings. Furthermore, to seek a more efficient network configuration, the intelligent search capabilities of the particle swarm optimization (PSO) algorithm have been utilized to systematically explore the optimal network structure parameters. The prediction results highlight the advantages of LSTM in handling time series data, particularly in predicting erosion rates with complex dynamic characteristics.

Multi-step ahead forecasting of electrical conductivity in rivers by using a hybrid Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) model enhanced by Boruta-XGBoost feature selection algorithm

Electrical conductivity (EC) is widely recognized as one of the most essential water quality metrics for predicting salinity and mineralization. In the current research, the EC of two Australian rivers (Albert River and Barratta Creek) was forecasted for up to 10 days using a novel deep learning algorithm (Convolutional Neural Network combined with Long Short-Term Memory Model, CNN-LSTM). The Boruta-XGBoost feature selection method was used to determine the significant inputs (time series lagged data) to the model. To compare the performance of Boruta-XGB-CNN-LSTM models, three machine learning approaches—multi-layer perceptron neural network (MLP), K-nearest neighbour (KNN), and extreme gradient boosting (XGBoost) were used. Different statistical metrics, such as correlation coefficient (R), root mean square error (RMSE), and mean absolute percentage error, were used to assess the models' performance. From 10 years of data in both rivers, 7 years (2012–2018) were used as a training set, and 3 years (2019–2021) were used for testing the models. Application of the Boruta-XGB-CNN-LSTM model in forecasting one day ahead of EC showed that in both stations, Boruta-XGB-CNN-LSTM can forecast the EC parameter better than other machine learning models for the test dataset (R = 0.9429, RMSE = 45.6896, MAPE = 5.9749 for Albert River, and R = 0.9215, RMSE = 43.8315, MAPE = 7.6029 for Barratta Creek). Considering the better performance of the Boruta-XGB-CNN-LSTM model in both rivers, this model was used to forecast 3–10 days ahead of EC. The results showed that the Boruta-XGB-CNN-LSTM model is very capable of forecasting the EC for the next 10 days. The results showed that by increasing the forecasting horizon from 3 to 10 days, the performance of the Boruta-XGB-CNN-LSTM model slightly decreased. The results of this study show that the Boruta-XGB-CNN-LSTM model can be used as a good soft computing method for accurately predicting how the EC will change in rivers.

Graph convolutional LSTM algorithm for real-time crash prediction on mountainous freeways

Accurate real-time traffic crash prediction is crucial for proactive traffic safety management. Currently, the majority of real-time models forecast crashes every 5 min to support different intelligent transportation systems. However, these intervals might be too short for practical use in manually implementing proactive traffic safety measures such as deploying traffic law enforcement and emergency rescue resources. Therefore, this study develops hourly crash prediction models to provide network operators with sufficient time to take measures in advance. A section of a mountainous freeway in Guizhou province is divided into homogeneous segments, with crash data, traffic operations data, and meteorological data being collected hourly. As the result is an imbalanced dataset of crash and non-crash instances, the training dataset is resampled using synthetic minority over-sampling technique (SMOTE) to address the issue. To fully capture the complex spatiotemporal relationships in the data and achieve high crash prediction accuracy, a graph convolutional network-long short-term memory (GCN-LSTM) model is constructed for the first time, combining a graph convolutional network (GCN) and long short-term memory (LSTM) neural network. For comparison purposes, LSTM, Extreme Gradient Boosting (XGBoost), and logistic regression (LR) models are developed. The results show that the GCN-LSTM model outperforms other models in hourly traffic crash prediction, and the optimal prediction performance is achieved with the crash-to-non-crash ratio of 1:4. The GCN-LSTM method is found to effectively capture the complex spatiotemporal relationships in prediction data and to handle imbalanced traffic crash data.

Multi-information based cloud–edge–end collaborative computational tasks offloading for industrial IoT systems

Cloud–edge–end collaborative computational task offloading (CEETO) is a promising method in industrial Internet-of-things (IIoT) to support massive computational tasks generated by equipment that has low energy/computation ability. In this work, we propose a new CEETO scheme by invoking the deep learning method (DL) with the aid of a multi-information analysis approach. Firstly, considering the delay constraints of the real-time tasks and the processing ability constraints of the cloud/edge/end servers, we formulate the CEETO problem to achieve the lowest system delay by establishing contact between CEETO problem and multiple information, such as the time-related locations and tasks requirements/features. Then, we tailor a long-short term memory network (LSTMN) to analyze the relation among time, locations and task requirements/features for predicting multiple information. Finally, the predicted multiple information is utilized for the final offloading strategy generation by invoking the simulated annealing algorithm (SAA). As the proposed CEETO process is invoked based on the predictions of multiple information, it is particularly suitable for the planning, scheduling and deployment of cloud–edge–end resources in massive equipment IIoT scenarios. Simulation results show that our proposed scheme can achieve effective computational task offloading.