Multi-scale sequential feature selection for disease classification using Raman spectroscopy data.

Raman spectroscopy (RS) optical technology promises non-destructive and fast application in medical disease diagnosis in a single step. However, achieving clinically relevant performance levels remains challenging due to the inability to search for significant Raman signals at different scales. Here we propose a multi-scale sequential feature selection method that can capture global sequential features and local peak features for disease classification using RS data. Specifically, we utilize the Long short-term memory network (LSTM) module to extract global sequential features in the Raman spectra, as it can capture long-term dependencies present in the Raman spectral sequences. Meanwhile, the attention mechanism is employed to select local peak features that were ignored before and are the key to distinguishing different diseases. Experimental results on three public and in-house datasets demonstrate the superiority of our model compared with state-of-the-art methods for RS classification. In particular, our model achieves an accuracy of 97.9 ± 0.2% on the COVID-19 dataset, 76.3 ± 0.4% on the H-IV dataset, and 96.8 ± 1.9% on the H-V dataset.

Assessment of humoral and cellular immunity after bivalent BNT162b2 vaccination and potential association with reactogenicity

Nowadays, new challenges around increasing production quality and productivity, and decreasing energy consumption, are growing in the manufacturing industry. In order to tackle these challenges, it is of vital importance to monitor the health of critical components. In the machine tool sector, one of the main aspects is to monitor the wear of the cutting tools, as it affects directly to the fulfillment of tolerances, production of scrap, energy consumption, etc. Besides, the prediction of the remaining useful life (RUL) of the cutting tools, which is related to their wear level, is gaining more importance in the field of predictive maintenance, being that prediction is a crucial point for an improvement of the quality of the cutting process. Unlike monitoring the current health of the cutting tools in real time, as tool wear diagnosis does, RUL prediction allows to know when the tool will end its useful life. This is a key factor since it allows optimizing the planning of maintenance strategies. Moreover, a substantial number of signals can be captured from machine tools, but not all of them perform as optimum predictors for tool RUL. Thus, this paper focuses on RUL and has two main objectives. First, to evaluate the optimum signals for RUL prediction, a substantial number of them were captured in a turning process and investigated by using recursive feature elimination (RFE). Second, the use of bidirectional recurrent neural networks (BRNN) as regressive models to predict the RUL of cutting tools in machining operations using the investigated optimum signals is investigated. The results are compared to traditional machine learning (ML) models and convolutional neural networks (CNN). The results show that among all the signals captured, the root mean squared (RMS) parameter of the forward force ( F-y ) is the optimum for RUL prediction. As well, the bidirectional long-short term memory (BiLSTM) and bidirectional gated recurrent units (BiGRU), which are two types of BRNN, along with the RMS of F-y signal, achieved the lowest root mean squared error (RMSE) for tool RUL, being also computationally the most demanding ones.

Prediction of atmospheric pollutants in urban environment based on coupled deep learning model and sensitivity analysis.

Accurate and efficient predictions of pollutants in the atmosphere provide a reliable basis for the scientific management of atmospheric pollution. This study develops a model that combines an attention mechanism, convolutional neural network (CNN), and long short-term memory (LSTM) unit to predict the O3 and PM2.5 levels in the atmosphere, as well as an air quality index (AQI). The prediction results given by the proposed model are compared with those from CNN-LSTM and LSTM models as well as random forest and support vector regression models. The proposed model achieves a correlation coefficient between the predicted and observed values of more than 0.90, outperforming the other four models. The model errors are also consistently lower when using the proposed approach. Sobol-based sensitivity analysis is applied to identify the variables that make the greatest contribution to the model prediction results. Taking the COVID-19 outbreak as the time boundary, we find some homology in the interactions among the pollutants and meteorological factors in the atmosphere during different periods. Solar irradiance is the most important factor for O3, CO is the most important factor for PM2.5, and particulate matter has the most significant effect on AQI. The key influencing factors are the same over the whole phase and before the COVID-19 outbreak, indicating that the impact of COVID-19 restrictions on AQI gradually stabilized. Removing variables that contribute the least to the prediction results without affecting the model prediction performance improves the modeling efficiency and reduces the computational costs.

A decision support system for selecting the most suitable machine learning in healthcare using user parameters and requirements

The application of machine learning in the medical field is still limited. The main reason behind the lack of use is the unavailability of an easy-to-use machine learning system that targets non-technical users. The objective of this paper is to propose an automated machine learning system to aid non-technical users. The proposed system provides the user with simple choices to provide suggestions to the system. The system uses the combination of the user's choices and performance evaluation to select the most suited model from available options. In this study, we employed the system on a Parkinson's disease dataset. The templates for support vector machine and random forest algorithms are provided to the system. Support vector machines and random forests were able to produce 80% and 75% accuracy, respectively. The system used performance parameters of the system and user choices to select the most suited models for each test case. The support vector machine was selected as the most suited model in three test cases, while random forest was selected as the most suited for one test case. The test cases also showed that the weighted time parameter impacted the results heavily.Copyright © 2022 The Author(s)

Comparison Analysis of LSTM and CNN Variants with Embedding Word Methods for Sentiment Analysis on Food Consumption Behavior

Lockdowns, working from home, staying at home, and physical distance are expected to significantly impact consumer attitudes and behaviors during the COVID-19 pandemic. During the implementation of the Movement Control Order, Malaysians' food preferences are already shifting away, influencing new consumption behavior. Since it has played a significant role in many areas of natural language, mainly using social media data from Twitter, there has been increased interest in sentiment analysis in recent years. However, research on the performance of various sentiment analysis methodologies such as n-gram ranges, lexicon techniques, deep learning, word embedding, and hybrid methods within this domain-specific sentiment is limited. This study evaluates several approaches to determine the best approach for tweets on food consumption behavior in Malaysia during the COVID-19 pandemic. This study combined unigram and bigram ranges with two lexicon-based techniques, TextBlob and VADER, and three deep learning classi-fiers, Long Short-Term Memory Network (LSTM), Convolutional Neural Networks (CNN), and their hybridization. Word2Vector and GloVe are two-word embedding approaches used by LSTM-CNN. The embedding GloVe on TextBlob approach with a combination of Unigram + Bigram [1,2] range produced the best results, with 85.79% accuracy and 85.30% F1-score. According to these findings, LSTM outperforms other classifiers because it achieves the highest scores for both performance metrics. The classification performance can be improved in future studies if the dataset is more evenly distributed across each positive and negative label. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Rate Insight: A Comparative Study on Different Machine Learning and Deep Learning Approaches for Product Review Rating Prediction in Bengali Language

In this contemporary era of digital marketing, ecommerce has emerged as one of the most preferred methods for day-to-day shopping. Ever since the COVID-19 pandemic, online shopping behavior has forever changed to less or no human-to-human interaction. As a result, it is getting more difficult for e-commerce enterprises to observe and evaluate market trends, particularly when done through consumer behavior analysis. To identify behavioral patterns and customer review-rating discrepancies, extensive analysis of product reviews is a substantial research field. Lack of benchmark corpora and language processing techniques, predicting review ratings in Bengali has become increasingly problematic. This paper thoroughly analyzes the approach to product review rating prediction for Bengali text reviews exploiting our own constructed dataset that was collected from an e-commerce website called DarazBD1. We acquired product reviews with labels known as ratings of five sentiment classes, from "1"to "5". It is noteworthy that we established a well-balanced dataset using our automated scraping system and a significant amount of time and effort is spent to maintain quality standards through the human annotation process. Exploration of multiple approaches to machine learning models such as logistic regression, random forest, multinomial naïve Bayes, and support vector machine, the best classification accuracy score of 78.63% is achieved by SVM. Subsequently, using Word2Vec, FastText, and GloVe embeddings with three deep neural network(DNN) architectures: CNN, Bi-LSTM, and a combination of CNN and Bi-LSTM, CNN+Bi-LSTM gave the highest accuracy score of 75.25% among the DNN architectures. © 2022 IEEE.

Deep learning-based hybrid clinical decision support system algorithm for COVID-19 diagnosis via PCR graphics and Thorax CT images, preliminary data

Introduction: The limited sensitivity of microbiological testing, challenges in radiological differential diagnosis, and expectations of quick and accurate diagnosis required developing clinical decision support systems (CDSS). We propose a new deep learning-based hybrid CDSS that combines the advantageous aspects of thorax computed tomography(CT) and reverse transcriptase-polymerase chain reaction(PCR) to overcome the weakness of each one. Method(s): We retrospectively constructed a database that contains CT images of healthy subjects and patients with COVID-19 pneumonia(CP), bacterial/viral pneumonia(BVP), interstitial lung diseases(ILD), and PCR data of patients who were tested positive and negative for SARS-CoV-2. A new 3D-convolutional neural network (3D-CNN) and long short-term memory network(LSTM) based CDSS is developed to perform accurate and robust detection of COVID19 using CT images and PCR data. Result(s): Performance results of the proposed models (Fig1) provide highly reliable diagnosis of COVID-19 with 93.2% and 99.7% AUC for CT and PCR data, respectively. Conclusion(s): Proposed CDSS with state-of-the-art deep learning methods provides similar performance compared to both radiologists in CT evaluation and microbiologists in PCR evaluation and can be safely used. We plan to develop a hybrid CDSS algorithm further, combining laboratory data with CT and PCR models.

Deep learning coupled model based on TCN-LSTM for particulate matter concentration prediction

In this study, we combined the Temporal Convolutional Network (TCN) model with the Long Short-Term Memory (LSTM) network model and applied it to prediction of atmospheric particulate matter (PM) concentrations. The study area is Xi'an City, Shaanxi Province, and the study period is from January 2015 to July 2022. During this period, Xi'an exceeded China's National Ambient Air Quality Grade Ⅰ standard for PM for up to 70% of the days. The prediction results of the TCN-LSTM model were compared with those of deep learning models (Convolutional Neural Network-LSTM, TCN, and LSTM) and machine learning models (Support Vector Regression and Random Forest). The R2 values of the TCN-LSTM model were all >0.88, indicating better performance than that of the other five models, and the errors of the TCN-LSTM model were all lower than those of the other five models. The results showed that high-accuracy PM predictions using deep learning models can improve air quality monitoring by compensating for problems in the environmental monitoring process such as pollutant monitoring errors caused by instrument failures. Additionally, sensitivity analysis helps to identify the key factors influencing the behavior of PM. A sensitivity analysis of PM for different periods of COVID-19 found that PM2.5 is more sensitive to O3, while PM10 is mainly influenced by PM2.5. The sensitivity analysis for the whole period showed that PM was closely related to CO. Removing variables that do not contribute to the model output based on the sensitivity analysis results improves modeling efficiency while reducing operating costs and improving environmental monitoring activities and management strategies. © 2023 Turkish National Committee for Air Pollution Research and Control

An efficient lung disease classification from X-ray images using hybrid Mask-RCNN and BiDLSTM

Lung diseases mainly affect the inner lining of the lungs causing complications in breathing, airway obstruction, and exhalation. Identifying lung diseases such as COVID-19, pneumonia, fibrosis, and tuberculosis at the earlier stage is a great challenge due to the availability of insufficient laboratory kits and image modalities. The rapid progression of the lung disease can be easily identified via Chest X-rays and this serves as a major boon for the terminally ill patients admitted to Intensive Care Units (ICU). To enhance the decision-making capability of the clinicians, a novel lung disease prediction framework is proposed using a hybrid bidirectional Long-Short-Term-Memory (BiDLSTM)-Mask Region-Based Convolutional Neural Network (Mask-RCNN) model. The Crystal algorithm is used to optimize the scalability and convergence issues in the Mask-RCNN model by hyperparameter tuning. The long-range dependencies for lung disease prediction are done using the BiDLSTM architecture which is connected to the fully connected layer of the Mask RCNN model. The efficiency of the proposed methodology is evaluated using three publicly accessible lung disease datasets namely the COVID-19 radiography dataset, Tuberculosis (TB) Chest X-ray Database, and National Institute of Health Chest X-ray Dataset which consists of the images of infected lung disease patients. The efficiency of the proposed technique is evaluated using different performance metrics such as Accuracy, Precision, Recall, F-measure, Specificity, confusion matrix, and sensitivity. The high accuracy obtained when comparing the proposed methodology with conventional techniques shows its efficiency of it in improving lung disease diagnosis. Copyright © 2022 Elsevier Ltd

Can financial stress be anticipated and explained? Uncovering the hidden pattern using EEMD-LSTM, EEMD-prophet, and XAI methodologies.

Global financial stress is a critical variable that reflects the ongoing state of several key macroeconomic indicators and financial markets. Predictive analytics of financial stress, nevertheless, has seen very little focus in literature as of now. Futuristic movements of stress in markets can be anticipated if the same can be predicted with a satisfactory level of precision. The current research resorts to two granular hybrid predictive frameworks to discover the inherent pattern of financial stress across several critical variables and geography. The predictive structure utilizes the Ensemble Empirical Mode Decomposition (EEMD) for granular time series decomposition. The Long Short-Term Memory Network (LSTM) and Facebook's Prophet algorithms are invoked on top of the decomposed components to scrupulously investigate the predictability of final stress variables regulated by the Office of Financial Research (OFR). A rigorous feature screening using the Boruta methodology has been utilized too. The findings of predictive exercises reveal that financial stress across assets and continents can be predicted accurately in short and long-run horizons even at the time of steep financial distress during the COVID-19 pandemic. The frameworks appear to be statistically significant at the expense of model interpretation. To resolve the issue, dedicated Explainable Artificial Intelligence (XAI) methods have been used to interpret the same. The immediate past information of financial stress indicators largely explains patterns in the long run, while short-run fluctuations can be tracked by closely monitoring several technical indicators.

The MET Project: Distinguishing Multisystem Inflammatory Syndrome in Children from Typhus Using Artificial Intelligence

Background. Multisystem inflammatory syndrome in children (MIS-C) following SARS-CoV-2 infection shares features with other inflammatory states, notably Kawasaki Disease. The rickettsial infection murine typhus is also in the differential for MIS-C in endemic areas. As the therapeutic approaches differ, it is essential to distinguish these disorders soon after presentation, well before confirmatory serologic testing results. Our objective was to develop an algorithm to accurately predict MIS-C versus typhus. Methods. Retrospective review extracted demographic, clinical, and laboratory features available within 6 hours of presentation for 133 MIS-C and 87 typhus patients. 33 features were broken into 44 inputs and passed through an attention module to compute importance. Inputs were then entered into machine learning algorithms as MIS-C or typhus. Patients were divided into training and test cohorts respecting proportions in the dataset. An equation was built to calculate the 'MET' (MIS-C versus endemic typhus) score. Results. MIS-C patients were younger (8.4 v 11.2 years, p< 0.0001) and the majority (71%) presented on day 4-6 of fever;most typhus patients (84%) presented with >=6 days (mean 4.9 v 7.3 days, p< 0.0001). Typhus patients were more likely to have rash (86% v 51%, p< 0.0001) and MIS-C patients red eyes (71% v 36%, p< 0.0001), other features were similar. MIS-C patients had higher C-reactive protein levels (17.7 v 9.8 mg/dL), procalcitonin (14.0 v 0.48 ng/mL), fibrinogen (558 v 394 mg/ dL) and neutrophil-to-lymphocyte ratio (12 v 3.5), all p< 0.0001, other parameters were similar. MIS-C patients were also more likely to have elevated troponin (0.48 v 0.01 ng/mL, p< 0.0001) and require intensive care (66% v 6%, p< 0.0001). A long short term memory network outperformed 6 other models (99% accuracy using all 33 elements). The MET score predicted MIS-C versus typhus with 90% accuracy using only 10 features (sensitivity 90%, specificity 90%). Conclusion. The clinical and laboratory similarities between typhus and MIS-C present challenges, but they can be reliably distinguished using artificial intelligence with as little as 10 features. Our ongoing interprofessional collaboration aims to make the MET score readily available to clinicians for use in patient encounters.

Accuracy and Performance Enhancement in E-Healthcare System Using Advanced Iot Based Model

The Internet of Things consists of physical things with sensors, computing power, and software that link to other devices or systems through the Internet or another communication network. Currently, IoT is being employed in health care research. Using a healthcare dataset, a trained network was used in this research to make judgments. The study uses the LSTM model for prediction in the present IoT system to produce a reliable and adaptable solution for healthcare. Connecting everything to the internet so people may live in a secure and comfy environment is the goal of the Internet of Things. Connecting everything in our environment is the main goal of the IoT. The internet might have an impact on computers that aren't directly connected to it. With the aid of the Internet of Things, patients who live distant from a healthcare centre are being helped. Researching the present IoT system's problems is essential in order to develop an effective, practical, and scalable solution for improving healthcare. There must be a strong basis for health-care apps to succeed. It has been anticipated that mankind would face increased health care issues in the future with the introduction of COVID 19. Healthcare solutions that are more trustworthy and precise will be needed, along with the capacity to treat health concerns remotely in a highly effective manner. Existing IoT-based studies are being analyzed for their effectiveness and accuracy by researchers doing in-depth analyses. The integration of LSTM-based intelligent approach to IoT system is being proposed as a method to increase the accuracy of the medical prescription and prescribing system. Research is focusing on improving accuracy and performance. Copyright © 2022, Anka Publishers. All rights reserved.

COMPARATIVE ANALYSIS OF EPIDEMIOLOGICAL, FORECASTING AND DEEP LEARNING MODELS FOR COVID19 SPREAD IN INDIA

Covid19 is affecting across many nations and most population of the world. As per WHO there are 270million confirmed with about 5.3 million fatalities as on December 15th, 2021. Many governments, organizations and local bodies have been applying various models in order to estimate the disease spread and appliede varied strategies to curb the spread. There are many models proposed by mathematicians and statisticians for the same. In the current work a comparison is done with mathematical disease spread models SIR, SIRD, classic time series forecasting modelARIMA, and artificial neural network models RNN, LSTM with Covid19 India data. The study investigates the effect of disease containment policies and vaccination drives for Covid19 data in the context of India using SIR Model. All the models are built for multiple time prediction windows starting from 5 days up to 45 days. The models are evaluated with MAE, MAPE and RMSE for multiple states and India level data. It is inferred that the prediction time of 5 days has best results for SIR model. The ARIMA model can predict withacceptable performance up to 30 days. RNN and LSTM models can predict for 5 days within acceptable performance. The best model that can predict longer durations and has good performance is ARIMA model. A detailed report on the model details and performance is the outcome of this study.

PANGO: Prediction Model Based on Clustering of Time Series for Traffic Flow to Venues

With the increasing number of venues in the city, it provides better services for people. However, the ever-changing flow of people has also brought challenges to the management of venues, especially under the current regular prevention and control measures for COVID-19. Therefore, it is extremely necessary to propose a suitable prediction model for pedestrian volume to venues. The timing characteristics of venue's traffic flow determine that the accuracy of the prediction results by applying classical prediction models is unsatisfactory. In order to resolve the problem, in this paper, a Long Short Term Memory network (LSTM) combined with clustering of time series named PANGO is proposed. In PANGO, the temporal clustering is proposed to solve the short-term dependence of traffic flow data, while the long-term cycle prediction model is applied to obtain the longterm cycle characteristics, so as to improve the accuracy of prediction. Finally, the results of multi-dimensional experiments show that the prediction accuracy of PANGO model is improved by 11.8% compared with the traditional LSTM model. © 2022 IEEE.

Forecasting COVID-19 infections in the Arabian Gulf region.

In this paper, an empirical analysis of linear state space models and long short-term memory neural networks is performed to compare the statistical performance of these models in predicting the spread of COVID-19 infections. Data on the pandemic daily infections from the Arabian Gulf countries from 2020/03/24 to 2021/05/20 are fitted to each model and a statistical analysis is conducted to assess their short-term prediction accuracy. The results show that state space model predictions are more accurate with notably smaller root mean square errors than the deep learning forecasting method. The results also indicate that the poorer forecast performance of long short-term memory neural networks occurs in particular when health surveillance data are characterized by high fluctuations of the daily infection records and frequent occurrences of abrupt changes. One important result of this study is the possible relationship between data complexity and forecast accuracy with different models as suggested in the entropy analysis. It is concluded that state space models perform better than long short-term memory networks with highly irregular and more complex surveillance data.

The Use of Arabic Language COVID-19 Tweets Analysis in IoT Applications

Social media platforms have become one of the most powerful tools for organizations and individuals to publish news and express thoughts or feelings. With the increasingly enormous number of internet users in Saudi Arabia, the need raised to analyze Arabic posts. Since the emergence of COVID-19 in the latest 2019, it lefts economies and businesses counting the cost while governments fight the spread of the virus with new compartmentalization measures. Keeping in view the importance of quick and timely data analysis and sharing for policy actions, Artificial intelligence (AI) has played a crucial role in facilitating the exchange of views and information between scientists and decision-makers during the Coronavirus pandemic, and they continue to do so. This work mined to these content-related tweets to see how people's feelings and expressions are changing. The results of this analysis can be used with integration with several IoT technologies to reduce the impact of covid-19 and drive new decisions in this field. For this goal, we proposed a Machine Learning (ML) models that can classify both of the sentiment and topic of Modern Standard Arabic (MSA) tweets and achieve high accuracy results. © 2021 IEEE.

Electric load forecasting based on Long-Short-Term-Memory network via simplex optimizer during COVID-19

Electric load forecasting is a challenging research, which is of great significance to the safe and stable operation of power grid in epidemic period. In this paper, Long-Short-Term-Memory (LSTM) model with simplex optimizer is proposed to forecast the electric load for an enterprise during the COVID-19 pandemic. The forecasting process consists of data processing, LSTM network construction and optimization. Firstly, some data processing steps includes information quantifying, electric load data cleaning, correlation-coefficient-based medical data filtering, clustering-based medical data and electric load data filling. Then LSTM-Based electric load forecasting model of enterprise is established during the COVID-19 pandemic. On this basis, LSTM network is trained and parameters are optimized via simplex optimizer. Finally, an example of the electric load forecasting of an enterprise during the COVID-19 pandemic is investigated. The forecasting results show that the reduced number of iterations is about 25% and the improved forecasting accuracy is about 5.6%. These results can be used as a reference for resuming production of enterprises and planning of electric grid.

Forecasting Daily Patient Arrivals during COVID-19 in Emergency Departments: A Deep Learning Approach

Study Objectives: As the fourth wave of coronavirus disease 2019 (COVID-19) surges in Michigan, most health care systems are experiencing an increased hospitalization rate of infected COVID-19 patients. Understanding the arrival rates of patients to the emergency department (ED) is fundamental in managing the limited health care resources. Our objective is to develop an accurate forecasting model based on ED patients’ arrival and COVID-19 status to help manage and facilitate a data-driven resource planning. Methods: A cohort study of patients with clinical suspicion of COVID-19 evaluated at 2 EDs within an integrated health system that cares for a racially diverse population. We included patient arrivals, COVID-19 status, and demographic information between the dates of January 1, 2020 and March 16, 2021. We developed deep learning models (Long Short-Term Memory (LSTM)) to forecast patient arrivals in two geographically diverse EDs (denoted as ED1 and ED2). We used data from January to December 2020 for model training and data from January 2021 to March 2021 for model validation. The models are evaluated based on the root mean squared error (RMSE), the square root of the average of the squared error between predicted and observed values, and the mean absolute error (MAE), which provides the mean absolute difference between the predicted and the observed ED patient arrival rates per day. Results: In ED1, there were 56, 61 total patient arrivals (1, 433 infected COVID-19 patients) with a mean age of 38.0 ± 21.2 years. A majority were female (33, 457, 59.1%) and 29, 040 (51.3%) were Black. The average patient arrival per day was 125.1 (SD 35.0) for those without COVID-19, and 3.3 (SD 3.6) for COVID-1 confirmed patients. In ED2, there were 74, 176 total patient arrivals (1, 546 infected COVID-19 patients) with a mean age of 45.0 ± 23.0 years. A majority were female (39, 521, 53.3%) and 10, 636 (14.3%) were Black. The average patient arrival per day was 164.7 (SD 33.2) for those without COVID-19, and 3.5 (SD 5.0) for COVID 19 confirmed patients. Figure 1 shows the observed and predicted patients’ arrival for the two EDs for regular and confirmed COVID-19 patients. The LSTM models show accurate prediction one week in advance of daily patient arrivals for ED1 and ED2 with RMSE scores of 17 and 20 patients, respectively. The MAE values imply that, on average, the forecast’s error from the true daily patient arrival rate is 13.9 and 16.0 for ED1 and ED2, respectively. For COVID-19 patient arrivals to ED1 and ED2, the RMSE score is 3 patients each, while th MAE values are 2.2 and 2.4, respectively. Conclusion: This study demonstrates that an average RMSE prediction score of 18.5 and 3 patient arrivals per day for regular and COVID-19 confirmed patients is possible across EDs using LSTM one week prior to forecasting. Future validation and implementation of such forecasting models could impact effective planning and allocation of limited ED and hospital resources. [Formula presented]

Investigation of recurrent networks LSTM in the problem of Covid-19 forecasting

in this paper problem of Covid-19 forecasting was considered and investigated. Review of different models and methods of pandemic forecasting are presented. For middle term forecasting indicators of Covid-19 the application of LSTM networks is suggested. The experimental investigations were carried out during which the optimal parameters LSTM network were found: sliding window size, forecasting interval and network architecture. The efficiency of LSTM in Covid-19 forecasting was estimated. © 2021 IEEE.

Investigating the Impact of the COVID-19 Pandemic on Crime Incidents Number in Different Cities

The COVID-19 pandemic is strongly affecting many aspects of human life and society around the world. To investigate whether this pandemic also influences crime, the differences of crime incidents numbers before and during the pandemic in four large cities (namely Washington DC, Chicago, New York City and Los Angeles) are investigated. Moreover, the Granger causal relationships between crime incidents numbers and new cases of COVID-19 are also examined. Based on that, new cases of COVID-19 with significant Granger causal correlations are used to improve the crime prediction performance. The results show that crime is generally impacted by the COVID-19 pandemic, but it varies in different cities and with different crime types. Most types of crimes have seen fewer incidents numbers during the pandemic than before. Several Granger causal correlations are found between the COVID-19 cases and crime incidents in these cities. More specifically, crime incidents numbers of theft in DC, Chicago and New York City, fraud in DC and Los Angeles, assault in Chicago and New York City, and robbery in Los Angeles and New York City, are significantly Granger caused by the new case of COVID-19. These results may be partially explained by the Routine Activity theory and Opportunity theory that people may prefer to stay at home to avoid being infected with COVID-19 during the pandemic, giving fewer chances for crimes. In addition, involving new cases of COVID-19 as a variable can slightly improve the performance of crime prediction in terms of some specific types of crime. This study is expected to obtain deeper insights to the relationships between the pandemic and crime in different cities, and to provide new attempts for crime prediction during the pandemic.