Data-driven machine learning quantifies ozone transport in the Hangzhou Bay urban cluster

The world currently is going through a serious pandemic due to the coronavirus disease (COVID-19). In this study, we investigate the gene structure similarity of coronavirus genomes isolated from COVID-19 patients, Severe Acute Respiratory Syndrome (SARS) patients and bats genes. We also explore the extent of similarity between their genome structures to find if the new coronavirus is similar to either of the other genome structures. Our experimental results show that there is 82.42% similarity between the CoV-2 genome structure and the bat genome structure. Moreover, we have used a bidirectional Gated Recurrent Unit (GRU) model as the deep learning technique and an improved variant of Recurrent Neural networks (i.e., Bidirectional Long Short Term Memory model) to classify the protein families of these genomes to isolate the prominent protein family accession. The accuracy of Gated Recurrent Unit (GRU) is 98% for labeled protein sequences against the protein families. By comparing the performance of the Gated Recurrent Unit (GRU) model with the Bidirectional Long Short Term Memory (Bi-LSTM) model results, we found that the GRU model is 1.6% more accurate than the Bi-LSTM model for our multiclass protein classification problem. Our experimental results would be further support medical research purposes in targeting the protein family similarity to better understand the coronavirus genomic structure.

Recently, accurate predicting outpatient visits is crucial in healthcare for optimizing service delivery and resource allocation. Time series models, particularly Deep Learning (DL) methods, have gained popularity in predicting demand and can potentially be used to predict demand for medical services, including outpatient hospital visits. This study aims to assess the potential of Bidirectional Long Short-Term Memory (Bi-LSTM) model in accurately predicting outpatient visits. The proposed model was tested with different set of parameters. Two important parameters adjusted in this study which are batch size and number of hidden nodes. The study also compares the performance of Bi-LSTM with other LSTM architectures, such as Vanilla LSTM and Stack LSTM. The results obtained show that Bi-LSTM performs best with batch size 64 and 10 hidden neurons. Besides, the results yielded indicate that the Bi-LSTM model performs exceptionally well with higher accuracy compared to other LSTM architectures. Overall, this study offers valuable insights into the use of Bi-LSTM models for predicting the number of outpatient visits.

In cybersecurity, the rise of fileless malware poses a significant challenge to endpoint security. Traditional detection methods often fail against these sophisticated attacks, necessitating advanced techniques like deep learning models. This study highlights the limitations of Bi-Directional Long Short-Term Memory (BLSTM) models in dynamic malware analysis and proposes enhancements through Convolutional Long Short-Term Memory (ConvLSTM) architecture. BLSTM models process input sequences in forward and backward directions, combining the results into one output. While this dual-layer approach improves analysis, it is time-consuming, potentially increasing the risk of fileless malware attacks. A key limitation of BLSTM is the lack of parameter sharing between forward and backward directions. This reduces its ability to capture spatial and temporal features simultaneously, hindering effectiveness in detecting fileless malware. To address this, the ConvLSTM model consolidates feature extraction within a single LSTM cell layer. ConvLSTM breaks down samples into subsequence and uses timesteps for additional feature extraction, enabling spatial-temporal data analysis and improving malware prediction accuracy. The model was tested using a dynamic malware dataset. Unlike traditional LSTM, ConvLSTM integrates convolutional layers, allowing parameter sharing across both spatial and temporal dimensions. This reduces computational complexity and improves model performance in handling multidimensional data. The research re-simulated prior work with BLSTM using the same malware dataset. The Spyder app ran the event simulator and the ConvLSTM model's results replaced BLSTM's using identical parameters. Time, accuracy and loss were the main performance metrics. ConvLSTM outperformed BLSTM, achieving 98% detection accuracy compared to BLSTM's 90%. It also significantly reduced processing time, averaging 10 seconds, while BLSTM took 22 seconds. ConvLSTM experienced lower losses, averaging 10% per epoch versus BLSTM's 20%. In conclusion, ConvLSTM offers superior performance over BLSTM in fileless malware detection. Its enhanced computational efficiency and ability to quickly mitigate threats make it a robust solution for fortifying endpoint security against evolving cyber threats. ConvLSTM holds potential in strengthening defence mechanisms against sophisticated malware attacks, providing a proactive approach to safeguarding networks and data.

To accommodate the rapid development of the distribution network of China, it is essential to research load forecasting methods with higher accuracy and stronger generalization capabilities in order to optimize distribution system control strategies, ensure the efficient and reliable operation of the power system, and provide a stable power supply to users. In this paper, a short-term load forecasting method is proposed for low-voltage distribution substations based on the bidirectional long short-term memory (BiLSTM) model. First, principal component analysis (PCA) and the fuzzy C-means method based on a genetic algorithm (GA-FCM) are used to extract the main influencing factors and classify different types of user electricity consumption behaviors. Then, the BiLSTM forecasting model utilizing the stochastic weight averaging (SWA) algorithm to enhance generalization capability is constructed. Finally, the load data from a low-voltage distribution substation in China over recent years are selected as a case study. Compared with conventional LSTM and BiLSTM prediction models, the annual electricity load curves for various user types forecasted by the PCA-BiLSTM model are more closely aligned with actual data curves. The proposed BiLSTM forecasting model exhibits higher accuracy and can forecast user electricity consumption data that more accurately reflect real-life usage.

The accurate prediction of Municipal Solid Waste (MSW) electricity generation is very important for the fine management of a city. This paper selects Shanghai as the research object, through the construction of a Bidirectional Long Short-Term Memory (BiLSTM) model, and chooses six influencing factors of MSW generation as the input indicators, to realize the effective prediction of MSW generation. Then, this study obtains the MSW electricity generation capacity in Shanghai by using the aforementioned prediction results and the calculation formula of theMSW electricity generation. The experimental results show that, firstly, the mean absolute error (MAE), mean absolute percentage error (MAPE), and root mean square error (RMSE) values of the BiLSTM model are 42.31, 7.390, and 63.32. Second, it is estimated that by 2025, the maximum and minimum production of MSW in Shanghai will be 17.35 million tons and 8.82 million tons under the three scenarios. Third, it is predicted that in 2025, the maximum and minimum electricity generation of Shanghai MSW under the three scenarios will be 512.752 GWh/y and 260.668 GWh/y. Finally, this paper can be used as a scientific information source for environmental sustainability decision-making for domestic MSW electricity generation technology.

Maize is one of the most important agricultural crops in the world which is affected by various pathogenetic diseases. These disease lead to low productivity and huge loss to the farmers. For this reason, detection of these disease in early stage by recognizing its symptomatic patterns will be beneficial for farmers. CNN based technique are widely used for classifying such symptoms which can detect all important features of an image. In this paper we have discussed a hybrid network by combining CNN with Bidirectional Long Short-Term Memory (BiLSTM) model is to detect and classify nine different disease of maize plant which are frequently affected diseases in this subcontinent. Here, BiLSTM has been used to create correlation among extracted features and to accelerate the recognition accuracy. For this reason, we have created a dataset with 29065 maize disease images where 80% sample were used for training and achieved accuracy 99.02%. This ensures that the model is very reliable for AI based disease recognition system and may contribute to increase the productivity of crops.

In humans, stress is a natural reaction to pressure, and when stress increases, the risk of mental health issues also increases. Misclassification can be caused by redundancy in certain physiological and behavioral features. To overcome this limitation, this study performs a Multivariate Analysis of Variance (MANOVA) based feature selection method, along with a Bidirectional Long Short-Term Memory (Bi-LSTM) model for efficient stress classification. The proposed MANOVA technique evaluates multiple dependent variables simultaneously, capturing correlations between physiological features to identify the most informative for the classification of mental stress. The Bi-LSTM model processes stress-related physiological signals, including heart rate and skin conductance, both forward and backward, effectively capturing long-term dependencies that help improve classification. Initially, ElectroCardioGram (ECG) signal data were obtained from two benchmark datasets. Then, label encoding techniques were employed for converting categorical features into numerical ones, and normalization was used to scale the data into a uniform range. The proposed stress classification model was experimentally evaluated on the WESAD and SWELL-KW datasets, achieving accuracies of 99.50% and 99.80%, respectively, outperforming existing approaches.

Ultra-short-term photovoltaic power prediction based on similar day clustering and temporal convolutional network with bidirectional long short-term memory model: A case study using DKASC data

Optimal deep transfer learning based ethnicity recognition on face images

This research details a method for predicting power usage that makes use of deep learning (DL) techniques, namely Bidirectional LSTM (BiLSTM) and Long Short-Term Memory (LSTM) models. For both the training and evaluation of the models, a real-world dataset was utilized, which included the hourly electricity usage of a Phoenix, USA, hospital building. Effective learning of temporal patterns was made possible by preprocessing, normalizing, and segmenting the data into sequences. Both LSTM and BiLSTM networks were developed and trained to perform 24-hour (short-term), 7-day (medium-term), and monthly (long-term) electricity consumption forecasting. A recursive multi-step prediction strategy was employed for extended forecasting horizons. They employed such measures of industry standards as MSE and Root RMSE to analyze prediction’s accuracy. Based on the findings, BiLSTM is superior to LSTM in the area of capturing complex consumption patterns, indicating that the former can be deployed to enhance energy control and optimization of smart HVAC systems and their energy management and planning

Sign language is the primary communication tool used by the deaf community and people with speaking difficulties, especially during emergencies. Numerous deep learning models have been proposed to solve the sign language recognition problem. Recently. Bidirectional LSTM (BLSTM) has been proposed and used in replacement of Long Short-Term Memory (LSTM) as it may improve learning long-team dependencies as well as increase the accuracy of the model. However, there needs to be more comparison for the performance of LSTM and BLSTM in LRCN model architecture in sign language interpretation applications. Therefore, this study focused on the dense analysis of the LRCN model, including 1) training the CNN from scratch and 2) modeling with pre-trained CNN, VGG-19, and ResNet50. Other than that, the ConvLSTM model, a special variant of LSTM designed for video input, has also been modeled and compared with the LRCN in representing emergency sign language recognition. Within LRCN variants, the performance of a small CNN network was compared with pre-trained VGG-19 and ResNet50V2. A dataset of emergency Indian Sign Language with eight classes is used to train the models. The model with the best performance is the VGG-19 + LSTM model, with a testing accuracy of 96.39%. Small LRCN networks, which are 5 CNN subunits + LSTM and 4 CNN subunits + BLSTM, have 95.18% testing accuracy. This performance is on par with our best-proposed model, VGG + LSTM. By incorporating bidirectional LSTM (BLSTM) into deep learning models, the ability to understand long-term dependencies can be improved. This can enhance accuracy in reading sign language, leading to more effective communication during emergencies.

Bi-LSTM based vertical total electron content prediction at low-latitude equatorial ionization anomaly region of South India

In the financial world, the forecasting of stock price gains significant attraction. For the growth of shareholders in a company's stock, stock price prediction has a great consideration to increase the interest of speculators for investing money to the company. The successful prediction of a stock's future cost could return noteworthy benefit. Different types of approaches are taken in forecasting stock trend in the previous years. In this research, a new stock price prediction framework is proposed utilizing two popular models; Recurrent Neural Network (RNN) model i.e. Long Short Term Memory (LSTM) model, and Bi-Directional Long Short Term Memory (BI-LSTM) model. From the simulation results, it can be noted that using these RNN models i.e. LSTM, and BI-LSTM with proper hyper-parameter tuning, our proposed scheme can forecast future stock trend with high accuracy. The RMSE for both LSTM and BI-LSTM model was measured by varying the number of epochs, hidden layers, dense layers, and different units used in hidden layers to find a better model that can be used to forecast future stock prices precisely. The assessments are conducted by utilizing a freely accessible dataset for stock markets having open, high, low, and closing prices.

<p>Stock price estimates are a complex task that requires a strong algorithm to calculate long-term prices. Stock prices are naturally related; hence it will be difficult to predict the cost. A proposed algorithm that uses market data to predict the share price using machine learning strategies such as a repetitive neural network called Long Short Term Memory, in which process weights are adjusted for each data point using a stochastic gradient. This program will provide better results compared to currently available pricing estimates algorithms. The network is trained and tested with a variety of input data to attract graphical results.</p>

Named entity recognition in electronic medical records is of great significance to the construction of medical knowledge maps. This paper proposes a model of bidirectional Long Short-Term Memory with a conditional random field layer(BiLSTM-CRF). In terms of simultaneously identifying 5 types of clinical entities from CCKS2018 Chinese EHRs corpus, the BiLSTM-CRF model finally achieved better performance than the baseline CRF model (F-score of 84.23% vs 82.49%).