Deep BiLSTM Research Articles

Deep Bi-LSTM with Binary Harris Hawkes Algorithm-Based Heart Risk Level Prediction

Deep Bi-LSTM with Binary Harris Hawkes Algorithm-Based Heart Risk Level Prediction

An Innovative Approach to Cardiovascular Disease Prediction: A Hybrid Deep Learning Model

The increasing prevalence of cardiovascular disorders has created an imperative need for accurate diagnoses. Despite the emergence of numerous techniques for disease classification and secure data transmission, a prevailing shortcoming is the lack of precision in decision-making. This study aimed to address this critical issue by introducing an innovative disease prediction model that uses a hybrid classifier. The proposed hybrid classifier combined deep Bidirectional Long-Short-Term Memory (deep Bi LSTM) and deep Convolutional Neural Network (deep CNN).To further improve its performance, the proposed approach employed hybridized swarm optimization to fine-tune fusion parameters and optimize the learning model for enhanced accuracy. This study focused on heart disease as its central concern, strengthening data security through the implementation of Diffi-Huffman based on Elliptic Curve Cryptography (ECC) during data transmission. The resulting automatic disease prediction model adopted the hybrid deep classifier, which was born from the amalgamation of two components: the interactive hunt-deep CNN classifier and the WoM-deep Bi LSTM. The proposed hybrid learning model achieved impressive accuracy, F-measure, sensitivity, and specificity of 97.716%, 97.848%, 98.021%, and 97.807%, respectively, marking a significant advance in the realm of cardiovascular disease prediction.

Estimation of walnut crop evapotranspiration under different micro-irrigation techniques in arid zones based on deep learning sequence models

To solve the problems of low accuracy of crop evapotranspiration (ETc) estimates in arid areas and to optimise ETc for precision irrigation in agriculture.Taking walnut in Wensu County, Aksu Region, Xinjiang as the research object, based on the multivariate time series data of walnut fertility, deep learning sequence models such as LSTM, GRU and BiLSTM were utilized to estimate the crop evapotranspiration (ETc) of walnut in arid zones under different micro-irrigation techniques. The results showed that the coefficient of determination (R2) of the LSTM, GRU and BiLSTM models were up to greater than 0.95 for both microirrigation techniques, and the BiLSTM model was relatively optimal. For the 3 g microirrigation technique, the R2, RMSEV and RPDV of the BiLSTM model were 0.978, 0.294 and 6.965 mm/d respectively. For the Hg microirrigation technique, the R2, RMSEV and RPDV of the BiLSTM model were 0.970, 0.339 and 6.026 mm/d respectively.With different scenarios of missing meteorological environmental variables, the BiLSTM model was used to estimate R2 > 0.98 for walnut crop evapotranspiration under different microirrigation techniques, with a range from 0.958 to 0. 986 under 3 g microirrigation technique, RMSEV ranged from 0.235 to 0.412 and RPDV ranged from 5.028 to 8.494. Under Hg microirrigation technique, R2 ranged from 0.942 to 0.979, RMSEV ranged from 0.288 to 0.477 and RPDV ranged from 4.253 to 6.905. Using the deep learning BiLSTM model, we can efficiently estimate the evapotranspiration of the walnut crop in arid zones with different micro-irrigation technologies. The evapotranspiration can provide a scientific theoretical reference for intelligent irrigation of walnut and agricultural water management.

Speech emotion recognition and classification using hybrid deep CNN and BiLSTM model

Speech emotion recognition and classification using hybrid deep CNN and BiLSTM model

Chinese Named Entity Recognition in Football Based on ALBERT-BiLSTM Model

Football is one of the most popular sports in the world, arousing a wide range of research topics related to its off- and on-the-pitch performance. The extraction of football entities from football news helps to construct sports frameworks, integrate sports resources, and timely capture the dynamics of the sports through visual text mining results, including the connections among football players, football clubs, and football competitions, and it is of great convenience to observe and analyze the developmental tendencies of football. Therefore, in this paper, we constructed a 1000,000-word Chinese corpus in the field of football and proposed a BiLSTM-based model for named entity recognition. The ALBERT-BiLSTM combination model of deep learning is used for entity extraction of football textual data. Based on the BiLSTM model, we introduced ALBERT as a pre-training model to extract character and enhance the generalization ability of word embedding vectors. We then compared the results of two different annotation schemes, BIO and BIOE, and two deep learning models, ALBERT-BiLSTM-CRF and ALBERT BiLSTM. It was verified that the BIOE tagging was superior than BIO, and the ALBERT-BiLSTM model was more suitable for football datasets. The precision, recall, and F-Score of the model were 85.4%, 83.47%, and 84.37%, correspondingly.

Human Activity Prediction Based on Forecasted IMU Activity Signals by Sequence-to-Sequence Deep Neural Networks.

Human Activity Recognition (HAR) has gained significant attention due to its broad range of applications, such as healthcare, industrial work safety, activity assistance, and driver monitoring. Most prior HAR systems are based on recorded sensor data (i.e., past information) recognizing human activities. In fact, HAR works based on future sensor data to predict human activities are rare. Human Activity Prediction (HAP) can benefit in multiple applications, such as fall detection or exercise routines, to prevent injuries. This work presents a novel HAP system based on forecasted activity data of Inertial Measurement Units (IMU). Our HAP system consists of a deep learning forecaster of IMU activity signals and a deep learning classifier to recognize future activities. Our deep learning forecaster model is based on a Sequence-to-Sequence structure with attention and positional encoding layers. Then, a pre-trained deep learning Bi-LSTM classifier is used to classify future activities based on the forecasted IMU data. We have tested our HAP system for five daily activities with two tri-axial IMU sensors. The forecasted signals show an average correlation of 91.6% to the actual measured signals of the five activities. The proposed HAP system achieves an average accuracy of 97.96% in predicting future activities.

Application of Bi-LSTM method for groundwater quality assessment through water quality indices

Water is a crucial resource in all economic activity, from farming to manufacturing. Water supplies are under much stress due to the ever-growing strain of the world population. Hence, effective water management is crucial to civilized society for raising living standards. To address the problems with drinking water quality, groundwater quality has to be frequently checked. Utilizing the water quality index, an effort has been made to comprehend groundwater quality (WQI). It is a technique for assessing water quality and a beneficial tool for determining how groundwater quality has changed over time and in different locations. Managing ecological systems and water resources can benefit greatly from machine learning approaches computation and considerable forecasting errors. Hence the deep neural network based long short-term memory network (LSTM), has been used for high performance. This study presents a deep learning-based Bi-LSTM model to predict the variables affecting groundwater quality. The suggested model's effectiveness was compared to several existing methods, including LSTM, RNN, and GRU. According to a comparative analysis, the suggested model has 0.98 % of accuracy and precision which exceeds all other approaches in terms of the best prediction performance, and it may serve as a decision-making basis for the comprehensive management of water quality.

A hybrid approach of Poisson distribution LDA with deep Siamese Bi-LSTM and GRU model for semantic similarity prediction for text data

A hybrid approach of Poisson distribution LDA with deep Siamese Bi-LSTM and GRU model for semantic similarity prediction for text data

Aspect-Based Sentiment Analysis of Customer Speech Data Using Deep Convolutional Neural Network and BiLSTM

Aspect-Based Sentiment Analysis of Customer Speech Data Using Deep Convolutional Neural Network and BiLSTM

WoM-based deep BiLSTM: smart disease prediction model using WoM-based deep BiLSTM classifier

WoM-based deep BiLSTM: smart disease prediction model using WoM-based deep BiLSTM classifier

A robust energy management approach in two-steps ahead using deep learning BiLSTM prediction model and type-2 fuzzy decision-making controller

A robust energy management approach in two-steps ahead using deep learning BiLSTM prediction model and type-2 fuzzy decision-making controller

A Deep BiLSTM Machine Learning Method for Flight Delay Prediction Classification

A Deep BiLSTM Machine Learning Method for Flight Delay Prediction Classification

Novel hybrid ARIMA–BiLSTM model for forecasting of rice blast disease outbreaks for sustainable rice production

In recent years, the application of artificial intelligence (AI) in agriculture has grown to be the most important research domain. The proposed work focuses on forecasting rice blast disease outbreaks in paddy crops. Disease management in the farm fields is the most difficult problem on the planet. There is a variety of reasons for this, first, a lack of farmers’ experience in diagnosing diseases, second experts’ experience in detecting diseases visually, and third unfavorable climate. In recent days, researchers have offered a variety of time-series techniques in different applications. This study adds time-series techniques to the field of agriculture by forecasting crucial rice blast disease outbreaks in the paddy crop of the Davangere region based on daily weather data obtained from KSNDMC. The statistical time-series technique called ARIMA is trained by employing real data of blast disease outbreaks in the Davangere region from the period of 2015–2019. Meanwhile, the deep BiLSTM model is trained by employing real weather data and blast disease outbreaks of the Davangere region. Both models are evaluated by performance metrics, such as mean squared error and mean absolute error. The proposed research is focused on the hybrid model ARIMA–BiLSTM which is a combination of the statistical ARIMA model and deep BiLSTM model. The seasonal component of the rice blast disease outbreak feature is extracted from the additive decompose function used in the ARIMA model and fed as a dependent feature for the BiLSTM model. According to the results obtained, the hybrid approach can successfully forecast blast disease outbreaks in paddy crops with a mean squared error of 0.037 and a mean absolute error of 0.028 compared to the statistical ARIMA and deep BiLSTM model.

Aspect based Nepali Text Sentiment Analysis using BiLSTM

Aspect-Based Sentiment Analysis is a text analysis technique that classifies the sentiment of a specific aspect in a text. With a rise in internet and social media access, people tend to decide to buy a product or to visit a restaurant based on comments and reviews. Categorical reviews can play a significant role in improving the business of a restaurant if a business entity can capture insights of those reviews and use productively. Since Nepali is a low resource language there is no significant numbers of the dataset for restaurant reviews. Therefore a dataset is created in the work to model a target aspect-based sentiment analysis in restaurant review domain. The dataset comprises of comments extracted from social media sites, websites, and manually prepared data. Data are cleaned, pre-processed then aspects and sentiments are extracted using POS tagging. Sentiments and target categories are classified using word embedding and the deep learning BiLSTM model. The sentiment classification model classifies the sentiments into positive or negative. The target classification model classifies text into 5 target categories Food (F), Price (P), Service (S), Hygiene (H), and Experience (E). The result was found to be good compared to similar models in other domains. The sentiment prediction model and target category prediction model attains the F1 Score of 0.88 and 0.76 respectively.

Personalized Deep Bi-LSTM RNN Based Model for Pain Intensity Classification Using EDA Signal.

Automatic pain intensity assessment from physiological signals has become an appealing approach, but it remains a largely unexplored research topic. Most studies have used machine learning approaches built on carefully designed features based on the domain knowledge available in the literature on the time series of physiological signals. However, a deep learning framework can automate the feature engineering step, enabling the model to directly deal with the raw input signals for real-time pain monitoring. We investigated a personalized Bidirectional Long short-term memory Recurrent Neural Networks (BiLSTM RNN), and an ensemble of BiLSTM RNN and Extreme Gradient Boosting Decision Trees (XGB) for four-category pain intensity classification. We recorded Electrodermal Activity (EDA) signals from 29 subjects during the cold pressor test. We decomposed EDA signals into tonic and phasic components and augmented them to original signals. The BiLSTM-XGB model outperformed the BiLSTM classification performance and achieved an average F1-score of 0.81 and an Area Under the Receiver Operating Characteristic curve (AUROC) of 0.93 over four pain states: no pain, low pain, medium pain, and high pain. We also explored a concatenation of the deep-learning feature representations and a set of fourteen knowledge-based features extracted from EDA signals. The XGB model trained on this fused feature set showed better performance than when it was trained on component feature sets individually. This study showed that deep learning could let us go beyond expert knowledge and benefit from the generated deep representations of physiological signals for pain assessment.

DAFA-BiLSTM: Deep Autoregression Feature Augmented Bidirectional LSTM network for time series prediction

Time series forecasting models that use the past information of exogenous or endogenous sequences to forecast future series play an important role in the real world because most real-world time series datasets are rich in time-dependent information. Most conventional prediction models for time series datasets are time-consuming and fraught with complex limitations because they usually fail to adequately exploit the latent spatial dependence between pairs of variables. As a successful variant of recurrent neural networks, the long short-term memory network (LSTM) has been demonstrated to have stronger nonlinear dynamics to store sequential data than traditional machine learning models. Nevertheless, the common shallow LSTM architecture has limited capacity to fully extract the transient characteristics of long interval sequential datasets. In this study, a novel deep autoregression feature augmented bidirectional LSTM network (DAFA-BiLSTM) is proposed as a new deep BiLSTM architecture for time series prediction. Initially, the input vectors are fed into a vector autoregression (VA) transformation module to represent the time-delayed linear and nonlinear properties of the input signals in an unsupervised way. Then, the learned nonlinear combination vectors of VA are progressively fed into different layers of BiLSTM and the output of the previous BiLSTM module is also concatenated with the time-delayed linear vectors of the VA as an augmented feature to form new additional input signals for the next adjacent BiLSTM layer. Extensive real-world time series applications are addressed to demonstrate the superiority and robustness of the proposed DAFA-BiLSTM. Comparative experimental results and statistical analysis show that the proposed DAFA-BiLSTM has good adaptive performance as well as robustness even in noisy environment.

ILoc-miRNA: extracellular/intracellular miRNA prediction using deep BiLSTM with attention mechanism.

The location of microRNAs (miRNAs) in cells determines their function in regulation activity. Studies have shown that miRNAs are stable in the extracellular environment that mediates cell-to-cell communication and are located in the intracellular region that responds to cellular stress and environmental stimuli. Though in situ detection techniques of miRNAs have made great contributions to the study of the localization and distribution of miRNAs, miRNA subcellular localization and their role are still in progress. Recently, some machine learning-based algorithms have been designed for miRNA subcellular location prediction, but their performance is still far from satisfactory. Here, we present a new data partitioning strategy that categorizes functionally similar locations for the precise and instructive prediction of miRNA subcellular location in Homo sapiens. To characterize the localization signals, we adopted one-hot encoding with post padding to represent the whole miRNA sequences, and proposed a deep bidirectional long short-term memory with the multi-head self-attention algorithm to model. The algorithm showed high selectivity in distinguishing extracellular miRNAs from intracellular miRNAs. Moreover, a series of motif analyses were performed to explore the mechanism of miRNA subcellular localization. To improve the convenience of the model, a user-friendly web server named iLoc-miRNA was established (http://iLoc-miRNA.lin-group.cn/).

Wading corvus optimization based text generation using deep CNN and BiLSTM classifiers

Many individuals suffer fromlocked-in syndrome or motor neuron disorders, leading to a loss of capability to control their muscles except eye movement. Many researchers have contributed to perform communication through blinking of eyes using classification. This research mainly concentrates on the prediction of texts from the EEG signal from the brain and suggestion of the next letters through BiLSTM classifier. The data is firstcollected from the repository and preprocessed using band pass filter inorder to overcome the interference of the signals. The bands are separated according to their frequencies and then statistical features, common spatial pattern, and quadruple symmetric pattern are extracted using feature extraction techniques. Finally, the text is predicted using the Wading Corvus optimization-based deep CNNclassifier, where the deep CNN classifier is optimally tuned using the wading Corvusoptimization. The suggestion of other letters based on the previous search ispredicted using the BiLSTM classifier. The efficiency of the Wading Corvus optimization-based deep CNN classifier is proved by measuring the parameters, such as accuracy, precision and recall, which are reported to be93.893%, 93.358%, and 95.571%, respectively. The supremacy of the BiLSTM classifier is proved by measuring the training loss,and the classifier attains a loss of 0.262, which is low compared with the existing state-of-art methods.

Development of lumbar spine MRI referrals vetting models using machine learning and deep learning algorithms: Comparison models vs healthcare professionals

IntroductionReferrals vetting is a necessary daily task to ensure the appropriateness of radiology referrals. Vetting requires extensive clinical knowledge and may challenge those responsible. This study aims to develop AI models to automate the vetting process and to compare their performance with healthcare professionals. Methods1020 lumbar spine MRI referrals were collected retrospectively from two Irish hospitals. Three expert MRI radiographers classified the referrals into indicated or not indicated for scanning based on iRefer guidelines. The reference label for each referral was assigned based on the majority voting. The corpus was divided into two datasets, one for the models' development with 920 referrals, and one included 100 referrals used as a held-out for the final comparison of the AI models versus national and international MRI radiographers. Three traditional models were developed: SVM, LR, RF, and two deep neural models, including CNN and Bi-LSTM. For the traditional models, four vectorisation techniques applied: BoW, bigrams, trigrams, and TF-IDF. A textual data augmentation technique was applied to investigate the influence of data augmentation on the models' performances. ResultsRF with BoW achieved the highest AUC reaching 0.99. CNN model outperformed Bi-LSTM with AUC = 0.98. With the augmented dataset, the performance significantly improved with an increase in F1 scores ranging from 1% to 7%. All models outperformed the national and international radiographers when compared on the hold-out dataset. ConclusionThe models assigned the referrals' appropriateness with higher accuracies than the national and international radiographers. Applying data augmentation significantly improved the models' performances. Implications for practiceThe outcomes suggest that the use of AI for checking referrals' eligibility could serve as a supporting tool to improve the referrals' management in radiology departments.

A New Stock Price Forecasting Method Using Active Deep Learning Approach

Stock price prediction is a significant research field due to its importance in terms of benefits for individuals, corporations, and governments. This research explores the application of the new approach to predict the adjusted closing price of a specific corporation. A new set of features is used to enhance the possibility of giving more accurate results with fewer losses by creating a six-feature set (that includes High, Low, Volume, Open, HiLo, OpSe), rather than the traditional four-feature set (High, Low, Volume, Open). The study also investigates the effect of data size by using datasets (Apple, ExxonMobil, Tesla, Snapchat) of different sizes to boost open innovation dynamics. The effect of the business sector in terms of the loss result is also considered. Finally, the study included six deep learning models, MLP, GRU, LSTM, Bi-LSTM, CNN, and CNN-LSTM, to predict the adjusted closing price of the stocks. The six variables used (High, Low, Open, Volume, HiLo, and OpSe) are evaluated according to the model’s outcome, showing fewer losses than the original approach, which utilizes the original feature set. The results show that LSTM-based models improved using the new approach, even though all models showed a comparative result wherein no model showed better results or continuously outperformed other models. Finally, the added new features positively affected the prediction models’ performance.