Stacked LSTM Research Articles

Automatic Defect Description of Railway Track Line Image Based on Dense Captioning.

The state monitoring of the railway track line is one of the important tasks to ensure the safety of the railway transportation system. While the defect recognition result, that is, the inspection report, is the main basis for the maintenance decision. Most previous attempts have proposed intelligent detection methods to achieve rapid and accurate inspection of the safety state of the railway track line. However, there are few investigations on the automatic generation of inspection reports. Fortunately, inspired by the recent advances and successes in dense captioning, such technologies can be investigated and used to generate textual information on the type, position, status, and interrelationship of the key components from the field images. To this end, based on the work of DenseCap, a railway track line image captioning model (RTLCap for short) is proposed, which replaces VGG16 with ResNet-50-FPN as the backbone of the model to extract more powerful image features. In addition, towards the problems of object occlusion and category imbalance in the field images, Soft-NMS and Focal Loss are applied in RTLCap to promote defect description performance. After that, to improve the image processing speed of RTLCap and reduce the complexity of the model, a reconstructed RTLCap model named Faster RTLCap is presented with the help of YOLOv3. In the encoder part, a multi-level regional feature localization, mapping, and fusion module (MFLMF) are proposed to extract regional features, and an SPP (Spatial Pyramid Pooling) layer is employed after MFLMF to reduce model parameters. As for the decoder part, a stacked LSTM is adopted as the language model for better language representation learning. Both quantitative and qualitative experimental results demonstrate the effectiveness of the proposed methods.

5G NOMA Defense Application Environment and Stacked LSTM Network Architectures

In 5G and beyond 5G wireless communication networks, the NOMA scheme is widely considered a major non-orthogonal access technique for improving system capacity and data rates. The main challenges in current NOMA systems are limited channel feedback and the difficulty of integrating it with advanced adaptive coding and modulation algorithms. This study analyses S-LSTM-based DL NOMA receivers in i.i.d. Nakagami-m fading channel circumstances as opposed to previously presented solutions. The LSTM has the advantage of responding dynamically to changing channel conditions. When compared to a typical NOMA system, a typical NOMA system has a 12% lower outage probability, a 39% increase in net throughput, and a maximum SER reduction of 48%. Complex modulated M-ary PSK and M-ary QAM data symbols are employed in D/L NOMA transmission. Classic receivers such as LS and MMSE are outperformed by the S-LSTM-based DL-NOMA receiver. The CP and non-linear clipping noise simulation curves compare the performance of the MMSE and LS receivers with that of the DL NOMA receiver in real-time propagation circumstances. The DL-based detector outperforms the MMSE for SNRs greater than 15 dB because the S-LSTM method is more robust than the clipping noise.

Advanced video anomaly detection using 2D CNN and stacked LSTM with deep active learning-based model

Around the world, the video surveillance system has gained wide acceptance and astonishing growth due to its broad applications. The surveillance system has become a paramount tool and benchmark for analyzing the harmony and safety of society. Anomaly detection and its associated applications play a key role in the integrity of the system. The aim of anomaly detection is to find rare and sparse occurrences of events from videos. Developing an accurate and time-efficient system i s s till remains c hallenging due to t he dynamic nature of a nomalies. The deep learning-based end-to-end system with full use of both spatial and temporal features from the input videos is proposed. The model combines the use of 2DCNN and Stacked LSTM to extract frame-level features through an anisotropic Gunnar Farneback Optical Flow algorithm. The system is evaluated on the benchmarked datasets namely UCSD Ped1 and UCSD Ped2, and it achieves an AUC of 95% and 94% respectively. The experimental results indicate that the proposed method is superior to state-of-the-art algorithms.

Predicting the Kijang Emas Bullion Price using LSTM Networks

This study investigates the potential of Deep Learning techniques, specifically LSTM networks, in forecasting Kijang Emas future value over a long period. Six LSTM models comprising of Simple LSTM, Bidirectional LSTM, and Stacked LSTM architecture were built and trained against a 15-year historical price data for Kijang Emas. The models’ performance was then measured against ARIMA (5,1,0) as a baseline reference and evaluated against the RAE, MSE and RMSE metric. The results revealed that LSTM networks models performed well in forecasting Kijang Emas price based on the test dataset where the average RMSE was between 49.9 to 50.3 while the Bidirectional LSTM was found to exhibit better performance as compared to the other LSTM models.

A Modern Way to Predict Alzheimer's Disease While Using Stacked LSTM and CNN-CUKOO

Alzheimer's disease is the main neurological disorder that has a significant impact on the lives of those who are affected. Deep learning is used to develop a one-of-a-kind Alzheimer's disease prediction method. The proposed model is applied to the provided EEG data. The goal is to examine the state of the brain and assess the changes that occur during the interracial period. The earlier prediction technique aided in the early detection of illness and treatment to save patients' lives. Feature learning and categorization are common components of an illness prediction module. The feature extractor is used to pass raw EEG data, which reduces computing complexity and execution time. SLSTM (stacked long short-term memory) is a recurrent neural network that can connect prior knowledge to the current task. After noticing that a patient's temporal data could be useful for forecasting illness progression, We propose an SLSTM-based prediction model. As a result, an SLSTM, a network with fully linked layers and activation layers, was built to track the temporal relationship between features and the next stage of Alzheimer's disease. We built and tested the CNN-SLSTM deep-learning model using existing time series data categorization and prediction techniques. The results of the experiments showed that raw EEG data, when combined with the CNN-SLSTM model, can accurately predict Alzheimer's disease with a low error rate, indicating a high level of confidence in our method.

Basic Activity Recognition from Wearable Sensors Using a Lightweight Deep Neural Network

The field of human activity recognition has undergone a great development, making its presence felt in various sectors such as healthcare and supervision. The identification of fundamental behaviours that occur regularly in our everyday lives can be extremely useful in the development of systems that aid the elderly, as well as opening the door to the detection of more complicated activities in a Smart home environment. Recently, the use of deep learning techniques allowed the extraction of features from sensor’s readings automatically, in a hierarchical way through non-linear transformations. In this study, we propose a deep learning model that can work with raw data without any pre-processing. Several human activities can be recognized by our stacked LSTM network. We demonstrate that our outcomes are comparable to or better than those obtained by traditional feature engineering approaches. Furthermore, our model is lightweight and can be applied on edge devices. Based on our expertise with two datasets, we obtained an accuracy of 97.15% on the UCI HAR dataset and 99% on WISDM dataset.

Traffic prediction using MSSBiLS with self‐attention model

AbstractDue to gathering and examining a tremendous amount of traffic stream data, traffic blockage is the most important issue in the intelligent transportation framework. The main aim of this research work is “to find the appropriate deep learning time series models, to predict the future traffic of Denmark Aarhus city.” In this article, multivariate sequential stacked bidirectional LSTM with self‐attention model (MSSBiLS‐SA) is proposed to prevent the long term dependencies and also to reduce the error rate. Based on the traffic state, the high, medium, and low range of traffic is predicted to give the best route to the users, and visualization is performed with the help of the Android application. The proposed MSSBiLS‐SA attains higher accuracy 94.56% and 95.36%, sensitivity 88.74% and 92.5%, specificity 93.5% and 94.5%, precision 94.23% and 95.5%, F‐score 94.5% and 94.5%, and lower time complexity 10.19% and 12.43% shows better performance when compared with the existing methods, such as prediction and classification of traffic location analysis using optimized graph convolutional recurrent neural network and prediction and classification of traffic location analysis based multilayer perceptron network using long short term memory. Finally, the proposed method predicts and classifies the traffic location very accurately.

Short-Term Solar Power Predicting Model Based on Multi-Step CNN Stacked LSTM Technique

Variability in solar irradiance has an impact on the stability of solar systems and the grid’s safety. With the decreasing cost of solar panels and recent advancements in energy conversion technology, precise solar energy forecasting is critical for energy system integration. Despite extensive research, there is still potential for advancement of solar irradiance prediction accuracy, especially global horizontal irradiance. Global Horizontal Irradiance (GHI) (unit: KWh/m2) and the Plane Of Array (POA) irradiance (unit: W/m2) were used as the forecasting objectives in this research, and a hybrid short-term solar irradiance prediction model called modified multi-step Convolutional Neural Network (CNN)-stacked Long-Short-Term-Memory network (LSTM) with drop-out was proposed. The real solar data from Sweihan Photovoltaic Independent Power Project in Abu Dhabi, UAE is preprocessed, and features were extracted using modified CNN layers. The output result from CNN is used to predict the targets using a stacked LSTM network and the efficiency is proved by comparing statistical performance measures in terms of Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), Mean Squared Error (MAE), and R2 scores, with other contemporary machine learning and deep-learning-based models. The proposed model offered the best RMSE and R2 values of 0.36 and 0.98 for solar irradiance prediction and 61.24 with R2 0.96 for POA prediction, which also showed better performance as compared to the published works in the literature.

Prediction of PM2.5 Concentration on the Basis of Multi-Time Scale Fusion

Long-term prediction of hour-concentration of PM2.5 (particles in atmospheric suspension with effective dimensions equal or lower than 2.5 microns) is of great significance for environmental protection and people’s health. At present, the prediction of hour-concentration of PM2.5 is mostly single-step prediction, which is to predict PM2.5 concentration at a future time point based on a period of historical data. In this paper, a model based on multi-time scale fusion is proposed to study single-step prediction and multi-step prediction, respectively. Experimental results show that the proposed model is better than stacked LSTM and CNN-LSTM in predicting PM2.5 hour-concentration.

A Novel Multi-Module Approach to Predict Crime Based on Multivariate Spatio-Temporal Data Using Attention and Sequential Fusion Model

Forecasting crime is complex since several complicated aspects contribute to a crime. Predicting crime becomes more challenging because of the enormous number of everyday crime episodes in varied places. Though there are many established machine learning and deep learning techniques, law enforcement officers face challenges preventing crime from occurring promptly. An efficient way of law enforcement is required to lower the crime rates. This paper proposed an effective multi-module method for predicting crime using deep learning techniques. Our proposed method has two modules: Feature Level Fusion and Decision Level Fusion. The first module employs temporal-based Attention LSTM, Spatio-Temporal based Stacked Bidirectional LSTM, and Fusion model. The Fusion model leverages the prior two model’s training data. The temporal-based model is the source model for the transfer learning technique on the dataset of different cities. By applying this technique, the training time of the model is reduced. In the second module, the Spatio-Temporal based Attention-LSTM, Stacked Bidirectional LSTM, and the result of feature-level fusion module are used to get the final prediction. The proposed architecture predicts the next hour based on the data from the past twenty-four hours. The estimated number of crimes in any category for a particular location can be obtained as the output of our suggested model. It also enables law enforcement to get insight into future crime occurrences based on category, time, and location. This work concentrated mainly on the USA’s San Francisco and Chicago cities for the experimental analysis. For the San Francisco and Chicago datasets, our model has the Mean Absolute Error of 0.008, 0.02, the Coefficient of Determination of 0.95 and 0.94, and the Symmetric Mean Absolute Percentage Error of 1.03% and 0.6%, respectively. The proposed model outperforms numerous other well-known models.

Prediction of solar cycle 25 using deep learning based long short-term memory forecasting technique

• Peak & timing of solar cycle 25 has been predicted using sunspot number data. • The concept of Deep Learning algorithm has been used for forecasting. • The proposed model showed good performance on various time intervals with low RMSE value. • Validation of model performance has been done by predicting past solar cycles 20–24. • The model captured non-linearity and cyclic trend within the data with good RMSE. In the current work we have used the deep learning based long short-term memory model to predict the strength and peak time of solar cycle 25 by employing the monthly smoothed sunspot number data obtained from WDC-SILSO, Royal Observatory of Belgium, Brussels. We have used the stacked LSTM forecasting model to predict the upcoming cycle 25. From our analysis it has been shown that our proposed model is capable of capturing long term dependencies as well as trend within the data. For cycle 20 and 21 the error difference between predicted as well as observed peak value is 2.3 and 0.7 respectively while the peak prediction error is 1.47% and 0.30%. The RMSE of the model for cycle 20 and 21 is 3.97 and 4.34 respectively. For cycle 22, the AE and RE is 4.6 and 2.16% while the RMSE of the model for this case is 4.50. The predicted peak amplitude of solar cycle 23 and 24 from our proposed model has a relative error of 1.75% and 1.99% respectively from the observed value while the RMSE is 3.4 for cycle 23 and 4.2 for cycle 24. Our projected prediction of cycle 25 using the proposed LSTM model, says that it will be stronger than cycle 24 and weaker than cycle 23. The solar cycle 25 will peak with an amplitude of sunspot number at 171.9 ± 3.4 and will be 47 % stronger than cycle 24. The solar cycle 25 will reach its peak in August 2023 ± 2 months.

Fault Diagnosis Based on Attention Collaborative LSTM Networks for NPC Three-Level Inverters

To address the problem, it is difficult to extract fault features for neutral-point-clamped (NPC) three-level inverters under nonstationary conditions. This article proposes a multi-information feature fusion diagnosis method based on attention collaborative stacked long short-term memory (ASLSTM) neural networks. First, parallel structural stacked LSTM networks are constructed, which are used to automatically extract features from multisource time-series data. Then, the attention mechanism is applied to weight these features adaptively. Finally, the fault is identified by features that integrate multiple sources of information. In addition, the quantum particle swarm optimization (QPSO) algorithm is used to intelligently tune the hyperparameters of ASLSTM to improve the reasonableness of hyperparameter selection for the diagnostic model. The simulation results of the fault diagnosis of the NPC three-level inverter circuit show that the proposed method is able to extract high-discriminative features from raw data and has better diagnostic results under various conditions compared with other schemes.

Reservoir Production Prediction Model Based on a Stacked LSTM Network and Transfer Learning.

Gas injection and water injection are common and effective methods to improve oil recovery. To ensure its production effect, it is necessary to simulate the oilfield production process. However, traditional composition simulation runs a large number of calculations and takes a long time. Through the analysis of relevant data, we found that production is affected by many factors and has a strong sequential character. Therefore, this paper proposes a deep learning model for reservoir production prediction based on stacked long short-term memory network (LSTM). It is applied to other well patterns with a short production time and a few samples in the same oilfield block by transfer learning. The model achieves an effective combination with the actual reservoir production process. At the same time, it uses the knowledge learned from the well pattern with sufficient historical data to assist in the establishment of the model of the well pattern with limited data. This can obtain accurate prediction results and save the model training time, thus getting more effective application effects than composition simulation. This paper verifies the effectiveness of the proposed method through the data and multiple different injection combinations of the Tarim oilfield.

Multivariate data driven prediction of COVID-19 dynamics: Towards new results with temperature, humidity and air quality data

Since the start of the COVID-19 pandemic many studies investigated the correlation between climate variables such as air quality, humidity and temperature and the lethality of COVID-19 around the world. In this work we investigate the use of climate variables, as additional features to train a data-driven multivariate forecast model to predict the short-term expected number of COVID-19 deaths in Brazilian states and major cities. The main idea is that by adding these climate features as inputs to the training of data-driven models, the predictive performance improves when compared to equivalent single input models. We use a Stacked LSTM as the network architecture for both the multivariate and univariate model. We compare both approaches by training forecast models for the COVID-19 deaths time series of the city of São Paulo. In addition, we present a previous analysis based on grouping K-means on AQI curves. The results produced will allow achieving the application of transfer learning, once a locality is eventually added to the task, regressing out using a model based on the cluster of similarities in the AQI curve. The experiments show that the best multivariate model is more skilled than the best standard data-driven univariate model that we could find, using as evaluation metrics the average fitting error, average forecast error, and the profile of the accumulated deaths for the forecast. These results show that by adding more useful features as input to a multivariate approach could further improve the quality of the prediction models.

A Convolutional Stacked Bidirectional LSTM with a Multiplicative Attention Mechanism for Aspect Category and Sentiment Detection

A Convolutional Stacked Bidirectional LSTM with a Multiplicative Attention Mechanism for Aspect Category and Sentiment Detection

Predicting the number of customer transactions using stacked LSTM recurrent neural networks

Predicting the number of customer transactions using stacked LSTM recurrent neural networks

Adaptive trading system based on LSTM neural network

With the continuous development of statistics and computer technology, the methods of technical analysis are also constantly expanding. A large number of statistical and mathematical methods have been applied to the analysis of stocks, which greatly expands the tools available for stock analysis. With the rapid development of deep learning and the great progress of computer hardware technology, statistical learning algorithms have been widely used in big data processing. Including convolutional neural network model and cyclic neural network model, have shown great advantages in the processing of time series such as images, speech and text. In today’s financial market, due to the existence of a large number of assets and fast information flow, people are more and more inclined to use quantitative investment. Quantitative investment as a means of financial sector, the New Deal, which is mainly composed of fundamental analysis and technical analysis from the traditional investment method, its core is to use mathematical model set up trading strategy, and with the help of computer technology on the financial data for quantitative analysis, and judgment, found that financial market rules, eliminating dependence on the experience of the people in the course of investment and emotional impact, so as to guide investment decisions. In order to improve the prediction ability of LSTM neural network model, this paper chooses the stack LSTM neural network model, and combines the stack LSTM neural network model with ADAM algorithm to achieve better prediction results. The empirical results show that the prediction effect of ADAM-based stack LSTM neural network is better than that of ADAM-based LSTM neural network.

기상정보 예측 성능 개선을 위한 빅데이터 병합 및 딥러닝 모델 최적화

급속한 산업화의 영향으로 많은 형태의 대기오염이 증가하고 있으며, 특히 미세먼지는 심장 및 폐 관련 질환을 발생시키거나 악화시키는 등 인체에 악영향을 끼친다. 본 연구에서는 미세먼지 피해를 줄이기 위한 해결방안으로 사전에 대응할 수 있도록 8시간 후의 미세먼지 수치를 예측한다. 우리는 정확한 예측을 위해 다양한 dataset을 구성하였고 시계열 데이터 예측에서 좋은 성능을 보여주는 LSTM 계열의 모델인 Stacked LSTM, LSTM AutoEncoder, LSTM Variational AutoEncoder를 이용한다. 연구 결과 데이터 수의 증가는 모델의 성능을 향상시키지 못했으나 feature 수의 증가는 모델의 성능을 향상시켰다. 또한, feature 수가 많은 dataset일수록 LSTM AutoEncoder가 가장 좋은 성능을 보였다.

Anomaly detection in cloud environment using artificial intelligence techniques

Now days the usage of cloud environment is rapidly increasing in all the fields to run applications in virtual machines instead of physical hardware based machine. This increases the service availability and also reduces the cost. The usage of openstack cloud environment is also increasing both in academics and industry as it provides open source cloud services to run the application both for research and for production environment. One of the challenges in cloud environment is that the detection and prediction of the anomalies before they occur. In the traditional approach, the anomalies are detected manually by keeping track of threshold level and heartbeat. The recent research is happening on using machine learning techniques in detecting the anomalies before they occur. In this paper, we propose a model for anomaly detection in openstack cloud environment. In the proposed model, we used Stacked and Bidirectional LSTM models to build the neural network. For the experiment the data is collected from openstack using collectd. The collected data sets 10 features and class label. Using LSTM neural network, we were able to detect the anomalies in openstack environment. The proposed model achieved the detection accuracy of 94.61% for training set and 93.98% for the test set using binary cross entropy function as a loss function.

E2LG: a multiscale ensemble of LSTM/GAN deep learning architecture for multistep-ahead cloud workload prediction

Efficient resource demand prediction and management are two main challenges for cloud service providers in order to control dynamic autoscaling and power consumption in recent years. The behavior of cloud workload time-series at subminute scale is highly chaotic and volatile; therefore, traditional machine learning-based time-series analysis approaches fail to obtain accurate predictions. In recent years, deep learning-based schemes are suggested to predict highly nonlinear cloud workloads, but sometimes they fail to obtain excellent prediction results. Hence, demands for more accurate prediction algorithm exist. In this paper, we address this issue by proposing a hybrid E2LG algorithm, which decomposes the cloud workload time-series into its constituent components in different frequency bands using empirical mode decomposition method which reduces the complexity and nonlinearity of prediction model in each frequency band. Also, a new state-of-the-art ensemble GAN/LSTM deep learning architecture is proposed to predict each sub band workload time-series individually, based on its degree of complexity and volatility. Our novel ensemble GAN/LSTM architecture, which employs stacked LSTM blocks as its generator and 1D ConvNets as discriminator, can exploit the long-term nonlinear dependencies of cloud workload time-series effectively specially in high-frequency, noise-like components. By validating our approach using extensive set of experiments with standard real cloud workload traces, we confirm that E2LG provides significant improvements in cloud workload prediction accuracy with respect to the mean absolute and standard deviation of the prediction error and outperforming traditional and state-of-the-art deep learning approaches. It improves the prediction accuracy at least 5% and 12% in average compared to the main contemporary approaches in recent papers such as hybrid methods which employs CNN, LSTM or SVR.