Forecasting of COVID-19 Pandemic Using ARIMA and Fb-Prophet Models: UK Case Study

This study aims to provide insights into predicting future cases of COVID-19 infection and rates of virus transmission in the UK by critically analyzing and visualizing historical COVID-19 data, so that healthcare providers can prepare ahead of time. In order to achieve this goal, the study invested in the existing studies and selected ARIMA and Fb-Prophet time series models as the methods to predict confirmed and death cases in the following year. In a comparison of both models using values of their evaluation metrics, root-mean-square error, mean absolute error and mean absolute percentage error show that ARIMA performs better than Fb-Prophet. The study also discusses the reasons for the dramatic spike in mortality and the large drop in deaths shown in the results, contributing to the literature on health analytics and COVID-19 by validating the results of related studies. Copyright © 2023 by SCITEPRESS - Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

Financial Volatility Modeling with the GARCH-MIDAS-LSTM Approach: The Effects of Economic Expectations, Geopolitical Risks and Industrial Production during COVID-19

Forecasting stock markets is an important challenge due to leptokurtic distributions with heavy tails due to uncertainties in markets, economies, and political fluctuations. To forecast the direction of stock markets, the inclusion of leading indicators to volatility models is highly important;however, such series are generally at different frequencies. The paper proposes the GARCH-MIDAS-LSTM model, a hybrid method that benefits from LSTM deep neural networks for forecast accuracy, and the GARCH-MIDAS model for the integration of effects of low-frequency variables in high-frequency stock market volatility modeling. The models are being tested for a forecast sample including the COVID-19 shut-down after the first official case period and the economic reopening period in in Borsa Istanbul stock market in Türkiye. For this sample, significant uncertainty existed regarding future economic expectations, and the period provided an interesting laboratory to test the forecast effectiveness of the proposed LSTM augmented model in addition to GARCH-MIDAS models, which included geopolitical risk, future economic expectations, trends, and cycle industrial production indices as low-frequency variables. The evidence suggests that stock market volatility is most effectively modeled with geopolitical risk, followed by industrial production, and a relatively lower performance is achieved by future economic expectations. These findings imply that increases in geopolitical risk enhance stock market volatility further, and that industrial production and future economic expectations work in the opposite direction. Most importantly, the forecast results suggest suitability of both the GARCH-MIDAS and GARCH-MIDAS-LSTM models, and with good forecasting capabilities. However, a comparison shows significant root mean squared error reduction with the novel GARCH-MIDAS-LSTM model over GARCH-MIDAS models. Percentage decline in root mean squared errors for forecasts are between 39% to 95% in LSTM augmented models depending on the type of economic indicator used. The proposed approach offers a key tool for investors and policymakers.

Performance of Bayesian Model Averaging (BMA) for Short-Term Prediction of PM10 Concentration in the Peninsular Malaysia

In preparation for the Fourth Industrial Revolution (IR 4.0) in Malaysia, the government envisions a path to environmental sustainability and an improvement in air quality. Air quality measurements were initiated in different backgrounds including urban, suburban, industrial and rural to detect any significant changes in air quality parameters. Due to the dynamic nature of the weather, geographical location and anthropogenic sources, many uncertainties must be considered when dealing with air pollution data. In recent years, the Bayesian approach to fitting statistical models has gained more popularity due to its alternative modelling strategy that accounted for uncertainties for all air quality parameters. Therefore, this study aims to evaluate the performance of Bayesian Model Averaging (BMA) in predicting the next-day PM10 concentration in Peninsular Malaysia. A case study utilized seventeen years' worth of air quality monitoring data from nine (9) monitoring stations located in Peninsular Malaysia, using eight air quality parameters, i.e., PM10, NO2, SO2, CO, O3, temperature, relative humidity and wind speed. The performances of the next-day PM10 prediction were calculated using five models' performance evaluators, namely Coefficient of Determination (R2), Index of Agreement (IA), Kling-Gupta efficiency (KGE), Mean Absolute Error (MAE), Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE). The BMA models indicate that relative humidity, wind speed and PM10 contributed the most to the prediction model for the majority of stations with (R2 = 0.752 at Pasir Gudang monitoring station), (R2 = 0.749 at Larkin monitoring station), (R2 = 0.703 at Kota Bharu monitoring station), (R2 = 0.696 at Kangar monitoring station) and (R2 = 0.692 at Jerantut monitoring station), respectively. Furthermore, the BMA models demonstrated a good prediction model performance, with IA ranging from 0.84 to 0.91, R2 ranging from 0.64 to 0.75 and KGE ranging from 0.61 to 0.74 for all monitoring stations. According to the results of the investigation, BMA should be utilised in research and forecasting operations pertaining to environmental issues such as air pollution. From this study, BMA is recommended as one of the prediction tools for forecasting air pollution concentration, especially particulate matter level.

Used Car Price Prediction in Surabaya Using Random Forest Regressor Algorithms

The use of private vehicles during the Covid-19 pandemic has increased because private vehicles, especially cars, are considered as the safest mode of transportation to maintain distance and prevent transmission of the Covid-19 virus. Based on data from two different Indonesian secondary car market place, a comparison of a price sample of Car X in the city of Surabaya with the specifications for the 2015 to 2018 car years with car milage under 1000 kilometers, the used cars have a variety of prices hence a used car price prediction system is needed so that people can find out the average price of used cars sold in the market. In this study the author will use the Random Forest Regressor as a machine learning algorithm to predict the price of a used car with a dataset from the AtapData website. The reason for choosing the Random Forest Regressor is because the algorithm has the power to handle large amounts of data with high dimensions with categorical and numerical data types. The evaluation method used in this study is the Root Mean Absolute Error which produces a value of 0.55612 for validation data and 0.56638 for testing data, while the evaluation proceed with Mean Absolute Error produces a value of 0.45208 for validation data and 0.47576 for testing data. © 2022 IEEE.

VAR, ARIMAX and ARIMA models for nowcasting unemployment rate in Ghana using Google trends

The analysis of the high volume of data spawned by web search engines on a daily basis allows scholars to scrutinize the relation between the user's search preferences and impending facts. This study can be used in a variety of economics contexts. The purpose of this study is to determine whether it is possible to anticipate the unemployment rate by examining behavior. The method uses a cross-correlation technique to combine data from Google Trends with the World Bank's unemployment rate. The Autoregressive Integrated Moving Average (ARIMA), Autoregressive Integrated Moving Average with eXogenous variables (ARIMAX) and Vector Autoregression (VAR) models for unemployment rate prediction are fit using the analyzed data. The models were assessed with the various evaluation metrics of mean absolute error (MAE), root mean square error (RMSE), mean absolute percentage error (MAPE), median absolute error (MedAE), and maximum error (ME). The average outcome of the various evaluation metrics proved the significant performance of the models. The ARIMA (MSE = 0.26, RMSE = 0.38, MAE = 0.30, MAPE = 7.07, MedAE = 0.25, ME = 0.77), ARIMAX (MSE = 0.22, RMSE = 0.25, MAE = 0.29, MAPE = 6.94, MedAE = 0.25, ME = 0.75), and VAR (MSE = 0.09, RMSE = 0.09, MAE = 0.20, MAPE = 4.65, MedAE = 0.20, ME = 0.42) achieved significant error margins. The outcome demonstrates that Google Trends estimators improved error reduction across the board when compared to model without them.

COVID-19 Outbreak Prediction by Using Machine Learning Algorithms

COVID-19 is a contagious disease and its several variants put under stress in all walks of life and economy as well. Early diagnosis of the virus is a crucial task to prevent the spread of the virus as it is a threat to life in the whole world. However, with the advancement of technology, the Internet of Things (IoT) and social IoT (SIoT), the versatile data produced by smart devices helped a lot in overcoming this lethal disease. Data mining is a technique that could be used for extracting useful information from massive data. In this study, we used five supervised ML strategies for creating a model to analyze and forecast the existence of COVID-19 using the Kaggle dataset” COVID-19 Symptoms and Presence.” RapidMiner Studio ML software was used to apply the Decision Tree (DT), Random Forest (RF), K-Nearest Neighbors (K-NNs) and Naive Bayes (NB), Integrated Decision Tree (ID3) algorithms. To develop the model, the performance of each model was tested using 10-fold cross-validation and compared to major accuracy measures, Cohan's kappa statistics, properly or mistakenly categorized cases and root means square error. The results demonstrate that DT outperforms other methods, with an accuracy of 98.42% and a root mean square error of 0.11. In the future, a devised model will be highly recommendable and supportive for early prediction/diagnosis of disease by providing different data sets. © 2023 Tech Science Press. All rights reserved.

Stock Market Prediction during COVID Using Stacked LSTM

In the field of computation, the art of predicting the stock market has always been a tough nut to crack for researchers. This is because stock prices are highly influential values. The prices depend on many factors, ranging from physical to physiological, rational and irrational, from geopolitical stability to the sentiments of the investors - all play a crucial role. Investors anticipate market conditions in the future for a successful investment. Hence considering the past stock prices as an embodiment of the factors mentioned above, we propose a stacked long-short-term-memory (LSTM) model to predict the closing index of stock prices during this highly uncertain pandemic period using root mean square error (RSME) as the performance indicator. The model is optimized to improve the prediction accuracy in order to achieve high performance stock forecasting. The dataset considered is from NIFTY 50 scaling across four sectors, namely - auto, bank, healthcare and metal from a duration of 30th January 2020 to 31st March 2022. This paper aims to consider the historical data to analyze future patterns and insights. © 2022 IEEE.

Forecasting the stock price of vaccine manufacturers in China Using Machine Learning and Econometrics Model

Due to the sudden outbreak of COVID-19, there is a high volatility in stock price of vaccine manufacturers in China (Between December 15, 2020 and December 13, 2021, average monthly volatility of these companies is 986). The aim of this paper is to compare the price prediction result of four algorithms: Multivariable Regression Model (MLR), Auto Regressive Integrated Moving Average Model (ARIMA), Back Propagation Neural Network Model (BP-NN), Random Forest Regression (RF), and decide which one has a better performance. Data from December 2020 to December 2021 is collected from Wind and the 8 stocks is selected in leading companies in vaccine industry. It turns out that BP-NN Model gives the best result in predicting stock price of vaccine manufacturers, measured using commonly used indicator, i.e., root-mean-square error (RMSE) and R Square (R2). So next time in the similar situation, BP-NN can be seen as a powerful tool to help us make decision. This paper would help investors build an optimal strategy in selecting stocks in terms of pharmaceutical industry. © 2022 SPIE.

Energy Consumption Prediction for Fused Deposition Modelling 3D Printing Using Machine Learning

Additive manufacturing (AM) technologies are growing more and more in the manufacturing industry;the increase in world energy consumption encourages the quantification and optimization of energy use in additive manufacturing processes. Orientation of the part to be printed is very important for reducing energy consumption. Our work focuses on defining the most appropriate direction for minimizing energy consumption. In this paper, twelve machine learning (ML) algorithms are applied to model energy consumption in the fused deposition modelling (FDM) process using a database of the FDM 3D printing of isovolumetric mechanical components. The adequate predicted model was selected using four performance criteria: mean absolute error (MAE), root mean squared error (RMSE), R-squared (R2), and explained variance score (EVS). It was clearly seen that the Gaussian process regressor (GPR) model estimates the energy consumption in FDM process with high accuracy: R2 > 99%, EVS > 99%, MAE < 3.89, and RMSE < 5.8.

Swin Transformer for COVID-19 Infection Percentage Estimation from CT-Scans

Coronavirus disease 2019 (COVID-19) is an infectious disease that has spread globally, disrupting the health care system and claiming millions of lives worldwide. Because of the high number of Covid-19 infections, it has been challenging for medical professionals to manage this crisis. Estimating the Covid-19 percentage can help medical staff categorize patients by severity and prioritize accordingly. With this approach, the intensive care unit (ICU) can free up resuscitation beds for the critical cases and provide other treatments for less severe cases to efficiently manage the healthcare system during a crisis. In this paper, we present a transformer-based method to estimate covid-19 infection percentage for monitoring the evolution of the patient state from computed tomography scans (CT-scans). We used a particular Transformer architecture called Swin Transformer as a backbone network to extract the feature from the CT slice and pass it through multi-layer perceptron (MLP) to obtain covid-19 infection percentage. We evaluated our approach on the covid-19 infection percentage estimation challenge dataset, annotated by two expert radiologists. The experimental results show that the proposed method achieves promising performance with a mean absolute error (MAE) of 4.5042, Pearson correlation coefficient (PC) of 0.9490, root mean square error (RMSE) of 8.0964 on the given Val set leaderboard and a MAE of 3.5569, PC of 0.8547 and RMSE of 7.5102 on the given Test set Leaderboard. These promising results demonstrate the high potential of Swin Transformer architecture for this image regression task of covid-19 infection percentage estimation from CT-scans. The source code of this project can be found at: https://github.com/suman560/Covid-19-infection-percentage-estimation. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Forecasting the COVID-19 Space-time Dynamics in Brazil with Convolutional Graph Neural Networks and Transport Modals

One of the major challenges imposed by the SARS-CoV-2 pandemic is the lack of pattern in which the virus spreads, making it difficult to create effective policies to prevent and tackle the pandemic. Several approaches have been proposed to understand the virus behavior and anticipate its infection and death curves at country ans state levels, thus supporting containment measures. Those initiatives generalize well for general extents and decisions, but they do not predict so well the trajectory of the virus through specific regions, such as municipalities, considering their distinct interconnection profiles. Specially in countries with continental dimensions, like Brazil, too general decisions imply that containment measures are applied either too soon or too late. This study presents a novel scalable alternative to forecast the numbers of case and death by SARS-CoV-2, according to the influence that certain regions exert on others. By exploiting a single-model architecture of graph convolutional networks with recurrent networks, our approach maps the main access routes to municipalities in Brazil using the modals of transport, and processes this information via neural network algorithms to forecast at the municipal level ans for the whole country. We compared the performance in forecasting the pandemic daily numbers with three baseline models using Mean Absolute Error (MAE), Symmetric Mean Absolute Percentage Error (sMAPE) and Normalized Root Mean Square Error (NRMSE) metrics, with the forecasting horizon varying from 1 to 25 days. Results show that the proposed model overcomes the baselines when considering the MAE and NRMSE (p ˂0.01), being specially suitable for forecasts from 14 to 24 days ahead. Author

Real-Time Temperature Distribution Monitoring in Chinese Solar Greenhouse Using Virtual LAN

The internal air temperature of Chinese solar greenhouse (CSG) has the problem of uneven spatial and temporal distribution. To determine temperature distribution at different locations, we designed a greenhouse temperature real-time monitoring system based on virtual local area network (VLAN) and estimate, including interpolation estimation module, data acquisition, and transmission module. The temperature data were obtained from 24 sensors, and the Ordinary Kriging algorithm estimated the temperature distribution of the whole plane according to the data. The results showed that the real-time temperature distribution monitoring method established was fast and robust. In addition, data validity rate for VLAN technology deployed for data transmission was 2.64% higher than that of cellular network technology. The following results are obtained by interpolation estimation of temperature data using gaussian model. The average relative error (ARE) of estimate, mean absolute error (MAE), root mean square error (RMSE), and determination coefficient (R2) were −0.12 °C, 0.42 °C, 0.56 °C, and 0.9964, respectively. After simple optimization of the number of sensors, the following conclusions are drawn. When the number of sensors were decreased to 12~16, MAE, RMSE, and R2 were 0.40~0.60 °C, 0.60~0.80 °C, and >0.99, respectively. Furthermore, temperature distribution in the greenhouse varied in the east–west and north–south directions and had strong regularity. The calculation speed of estimate interpolation algorithm was 50~150 ms, and greenhouse Temperature Distribution Real-time Monitoring System (TDRMS) realized simultaneous acquisition, processing, and fast estimate.

Forecasting of Automobile Sales Based on Support Vector Regression Optimized by the Grey Wolf Optimizer Algorithm

With the development of the Internet and big data, more and more consumer behavior data are used in different forecasting problems, which greatly improve the performance of prediction. As the main travel tool, the sales of automobiles will change with the variations of the market and the external environment. Accurate prediction of automobile sales can not only help the dealers adjust their marketing plans dynamically but can also help the economy and the transportation sector make policy decisions. The automobile is a product with high value and high involvement, and its purchase decision can be affected by its own attributes, economy, policy and other factors. Furthermore, the sample data have the characteristics of various sources, great complexity and large volatility. Therefore, this paper uses the Support Vector Regression (SVR) model, which has global optimization, a simple structure, and strong generalization abilities and is suitable for multi-dimensional, small sample data to predict the monthly sales of automobiles. In addition, the parameters are optimized by the Grey Wolf Optimizer (GWO) algorithm to improve the prediction accuracy. First, the grey correlation analysis method is used to analyze and determine the factors that affect automobile sales. Second, it is used to build the GWO-SVR automobile sales prediction model. Third, the experimental analysis is carried out by using the data from Suteng and Kaluola in the Chinese car segment, and the proposed model is compared with the other four commonly used methods. The results show that the GWO-SVR model has the best performance of mean absolute percentage error (MAPE) and root mean square error (RMSE). Finally, some management implications are put forward for reference.

Hybrid LSTM-ARMA Demand-Forecasting Model Based on Error Compensation for Integrated Circuit Tray Manufacturing

Demand forecasting plays a crucial role in a company’s operating costs. Excessive inventory can increase costs and unnecessary waste can be reduced if managers plan for uncertain future demand and determine the most favorable decisions. Managers are demanding increasing accuracy in forecasting as technology advances. Most of the literature discusses forecasting results’ inaccuracy by suspending the model and reloading the data for model retraining and correction, which is extensively employed but causes a bottleneck in practice since users do not have the sufficient ability to correct the model. This study proposes an error compensation mechanism and uses the individuals and moving-range (I-MR) control chart to evaluate the requirement for compensation to solve the current bottleneck using forecasting models. The approach is validated using the case companies’ historical data, and the model is developed using a rolling long short-term memory (LSTM) to output the predicted values;then, five indicators are proposed for screening to determine the prediction statistics to be subsequently employed. Root mean squared error (RMSE), mean absolute percentage error (MAPE), and mean absolute error (MAE) compare the LSTM, rolling LSTM combined index, and LSTM-autoregressive moving average (ARMA) models. The results demonstrate that the RMSE, MAPE, and MAE of LSTM-ARMA are smaller than those of the other two models, indicating that the error compensation mechanism that is proposed in this study can enhance the prediction’s accuracy.

ANALYZE THE SPATIAL DISTRIBUTION OF DELIVERY MOTORCYCLE CRASHES AND IDENTIFY THE RELATED FACTORS

During COVID-19, the suspension of the dine-in option at restaurants had significantly increased online food delivery crashes in Taiwan. Nevertheless, the majority of current studies remain focused on the common motorcycle, which has distinct driving habits and routes than a delivery motorcycle. Even though some recent studies identified the variables contributing to delivery motorcycle crashes, they still restricted in defining crash severity model and did not account for spatial dependences. In this study, two different models were used in this study: the generalized linear model (GLM), and the geographically weighted negative binomial model (GWNBR) to estimate crash frequency in a non-stationary pattern. In 2020, there were 2314 delivery motorcycle crashes in Taipei, according to the study area. Besides that, the point of interests data from 456 villages in Taipei city was considered as related crash factors for further analysis. According to the results, GWNBR showed the best performance in terms of log-likelihood, Akaike Information Criterion (AIC), and Root Mean Square Error (RMSE). Furthermore, this research reveals that commercial areas and bus stations had a significant impact on delivery motorcycle crashes. As per the coefficient distribution, the effect is exacerbated in rural areas where the traffic policy is still a major concern. As the popularity of delivery food services grows, this topic will become even more important in the future.

Experimental Performance Analysis of Wi-SUN Channel Modelling Applied to Smart Grid Applications

The grid operation and communication network are essential for smart grids (SG). Wi-SUN channel modelling is used to evaluate the performance of Wi-SUN smart grid networks, especially in the last-mile communication. In this article, the distribution approximation of the received signal strength for IEEE 802.15.4g Wi-SUN smart grid networks was investigated by using the Rician distribution curve fitting with the accuracy improvement by the biased approximation methodology. Specifically, the Rician distribution curve fitting was applied to the received signal strength indicator (RSSI) measurement data. With the biased approximation method, the Rician K-factor, a non-centrality parameter (rs), and a scale parameter (σ) are optimized such that the lower value of the root-mean squared error (RMSE) is acheived. The environments for data collection are selected for representing the location of the data concentrator unit (DCU) and the smart meter installation in the residential area. In summary, the experimental results with the channel model parameters are expanded to the whole range of Wi-SUN’s frequency bands and data rates, including 433.92, 443, 448, 923, and 2440 MHz, which are essential for the successful data communication in multiple frequency bands. The biased distribution approximation models have improved the accuracy of the conventional model, by which the root mean-squared error (RMSE) is reduced in the percentage range of 0.47–3.827%. The proposed channel models could be applied to the Wi-SUN channel simulation, smart meter installation, and planning in smart grid networks.

Forecasting of Individuals' Movement in Society After Corona Pandemic Using RNNs with LSTM and GRUs

The health crisis that attributed to the quick spread of the COVID-19 has impacted the globe negatively in terms of economy, education, and transport and led to the global lockdown. The risk of the COVID-19 infection has been increased due to a lack of a successful cure for the disease. Thus, social distancing is considered the most appropriate precaution measure to control the viral spread throughout the world. In this study, a model was proposed for deep learning capable of predicting the movement of people in the pandemic in the short term (one day) to take precautions and control the COVID-19 infection. The proposed model consists of four phases: data collection, pre-processing phase, prediction stage, and evaluation and Comparison phase. The dataset is obtained from 428 mobility reports, collected based on data from users that have been selected for their Google Account location history for a country such as Iraq for 428 days. A deep learning algorithm such as Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and hybrid model (GRU&LSTM) is applied to pre-processed data to predict the movement of people. They are compared using statistical measures: Mean absolute error (MAE) and root mean square error (RMSE) for performance measurement of these machine learning algorithms. The results of the GRU are the sum of MAE 0.4277 and sum of RMSE 0.6470 for predict person path and movement with training time equal to 33.189 sec, while the results of the hybrid model are the sum of MAE 0.4355 and sum of RMSE 0.6563 for prediction and the training time equal to 53.144 sec, and the results of the LSTM are the sum of MAE 0.4395 and sum of RMSE 0.6612 for prediction and the training time equal to 100.752 sec. These statistical measurement values indicate proposed model GRU outperformed all other models, it showed a solid performance to predict person path and movement in coronavirus pandemic and took little time to train compared to other algorithms, while the hybrid algorithm showed good performance and a short period in training compared with the LSTM model. © 2021 IEEE.

Analysis of Prediction Data for the Third Wave of COVID-19 in Bogor Regency

The COVID-19 pandemic is far from over. The government has carried out several policies to suppress the development of COVID-19 is no exception in Bogor Regency. However, the public still has to be vigilant especially now we will face a year-end holiday that can certainly be a trigger for the third wave of COVID-19. Therefore, researchers aim to make predictions of the increase in positive cases, especially in the Bogor Regency area to help the government in making policies related to COVID-19. The algorithms used are Gaussian Process, Linear Regression, and Random Forest. Each Algorithm is used to predict the total number of COVID-19 cases for the next 21 days. Researchers approached the Time Series Forecasting model using datasets taken from the COVID-19 Information Center Coordinationn Center website. The results obtained in this study, the method that has the highest probability of accurate and appropriate data contained in the Gaussian Process method. Prediction data on the Linear Regression method has accurate results with actual data that occur with Root Mean Square Error 1202.6262. © 2021 IEEE.

IgG-IgM antibodies based infection time detection of COVID-19 using machine learning models

Over the last two years, most scientists have been researching the solution to the pandemic coronavirus disease 2019 (COVID-19). So, the effective inspection and the rapid diagnosis of COVID-19 provide a mitigation ability to the burden on healthcare systems. These research works focus on detecting and knowing the history of infection in terms of time and developed symptoms. In infections detection, artificial intelligence (AI) technologies increase the accuracy and efficiency of the adopted detection methods. These methods will aid the medical staff in classifying patients, essentially when there is a healthcare resources shortage. This paper proposed machine learning-based models for detecting the time of COVID-19 infection in weeks using the laboratory factors of detected antibodies immunoglobulins G and immunoglobulins M (IgG-IgM). This test is common and helpful in diagnosing the suspected patients who held a negative result for the reverse transcription-polymerase chain reaction (RT-PCR) test. The proposed models consider two machine learning models adopting root mean square error (RMSE) and mean absolute error (MAE) factors. The results show acceptable efficiency of performance that ranges from 80% to 100% for pointing the patient in any week of infection, to reduce the likelihood of transmitting the infection from patients who have developed symptoms but with false-negative RT-PCR test.

Video-Assisted Clinical Care for Remote Management of COVID-19

Due to the COVID-19 pandemic, the need for biomedical monitoring devices has increased for collecting vital signs from patients virtually. This paper will discuss developing an iOS mobile application that will ingest video frames in real-time to provide oxygen saturation and heart rate values. We provide two techniques for capturing data using 1) face monitoring under natural light and 2) fingertip monitoring using the iPhone’s flashlight. We observed an average Root Mean Squared Error (RMSE) of 3.6 for heart rate estimation using fingertip recordings and 13.25 using face recordings. For oxygen saturation, we obtained the RMSE for each class between SpO2 values 94 - 99%. The lowest RMSE provided from our application was 0.26 for fingertip recording and 0.22 for face recording at SpO2 level 96%. The highest RMSE was 6.34 for fingertip recording and 6.56 for face recording at SpO2 level 99%. These preliminary models will be further enhanced through a clinical study with UC Davis Health as we collect data from participants with respiratory diseases.