Multi-step CNN forecasting for COVID-19 multivariate time-series

The new coronavirus (COVID-19) has spread to over 200 countries, with over 36 million confirmed cases as of October 10, 2020. As a result, numerous machine learning models capable of forecasting the epidemic worldwide have been produced. This paper reviews and summarizes the most relevant machine learning forecasting models for COVID-19. The dataset is derived from the world health organization (WHO) COVID-19 dashboard, and it contains official daily counts of COVID-19 cases, fatalities, and vaccination use reported by countries, territories, and regions. We propose various convolutional neural network (CNN) based models such as CNN, single exponential smoothing CNN (S-CNN), moving average CNN (MA-CNN), smoothed moving average CNN (SMA-CNN), and moving average smoothed CNN (MAS-CNN). Here, MAPE and MSE are used to assess the suggested models. MAPE is frequently used to compare accuracy across time series with different scales. MSE, the model must strive for a total forecast equal to the entire demand. That is, optimizing MSE seeks to create a forecast that is right on average and so unbiased. The final result shows that SMA-CNN outperformed its baselines in both MAPE and MSE. The main contribution of this novel forecasting approach is a more accurate result as a base of the strategy of preventing COVID-19 spreads. © 2023, Universitas Ahmad Dahlan. All rights reserved.

Coping with COVID-19 shocks in rural Nepal

We examine shocks experienced by rural Nepali households during the COVID-19 pandemic. Households primarily experienced income and price shocks during a government-imposed lockdown. During this time, households managed to effectively protect consumption, and mostly relied on credit (26%), asset sales (10%) and savings (8%). Debt levels nearly doubled, with limited changes to savings. We then leverage a long-term randomized control trial (RCT) to assess whether beneficiaries of a livestock livelihood program are more resilient. Program beneficiaries are 6 percentage points less likely to take out new loans.

Robust optimal estimation of location from discretely sampled functional data

Estimating location is a central problem in functional data analysis, yet most current estimation procedures either unrealistically assume completely observed trajectories or lack robustness with respect to the many kinds of anomalies one can encounter in the functional setting. To remedy these deficiencies we introduce the first class of optimal robust location estimators based on discretely sampled functional data. The proposed method is based on M‐type smoothing spline estimation with repeated measurements and is suitable for both commonly and independently observed trajectories that are subject to measurement error. We show that under suitable assumptions the proposed family of estimators is minimax rate optimal both for commonly and independently observed trajectories and we illustrate its highly competitive performance and practical usefulness in a Monte‐Carlo study and a real‐data example involving recent Covid‐19 data. [ FROM AUTHOR] Copyright of Scandinavian Journal of Statistics is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full . (Copyright applies to all s.)

Mapping Life Expectancy Loss in Barcelona in 2020

We use a Bayesian spatio-temporal model, first to smooth small-area initial life expectancy estimates in Barcelona for 2020, and second to predict what small-area life expectancy would have been in 2020 in absence of covid-19 using mortality data from 2007 to 2019. This allows us to estimate and map the small-area life expectancy loss, which can be used to assess how the impact of covid-19 varies spatially, and to explore whether that loss relates to underlying factors, such as population density, educational level, or proportion of older individuals living alone. We find that the small-area life expectancy loss for men and for women have similar distributions, and are spatially uncorrelated but positively correlated with population density and among themselves. On average, we estimate that the life expectancy loss in Barcelona in 2020 was of 2.01 years for men, falling back to 2011 levels, and of 2.11 years for women, falling back to 2006 levels.

When should grocery stores adopt time-based pricing? Impact of competition and negative congestion externality

Consumers dread shopping during peak hours, and the Covid-19 pandemic has created additional safety concerns about overcrowding in addition to long waiting times. In view of consumer's congestion aversion, should competitive brick-and-mortar grocery stores charge higher prices during congested peak hours to smooth demand? To examine "whether and when” stores should adopt intraday time-based pricing under competition, we examine a 2-stage dynamic duopoly game. At the beginning of each stage, each store can make an irreversible decision to adopt time-based pricing by setting the peak-hour and normal-hour prices. We also endogenize consumer's shopping decisions (i.e., when and which store to shop) by incorporating the issue of negative congestion externality. Our equilibrium analysis reveals that time-based pricing is always beneficial for the stores, and both stores would adopt it eventually in equilibrium. As such, only two equilibria can sustain: either both firms adopt time-based pricing immediately in stage 1, or only one firm adopts in stage 1 while the other postpones its adoption until stage 2. Interestingly, due to the competitive dynamics, it is less likely for both firms to adopt immediately when consumers are more averse to congestion. Moreover, although the adoption of time-based pricing leads to differentiated price competition, it can "soften” price competition, causing both peak-hour and normal-hour prices to rise above the status quo equilibrium uniform prices. We find that time-based pricing can always induce demand smoothing and reduce congestion. Although time-based pricing creates value for the stores (through higher prices), it offers no benefit to consumers. © 2023 Production and Operations Management Society.

Stock Price Forecasting of IBEX35 Companies in the Petroleum, Electricity, and Gas Industries

In recent years, time series forecasting has become an essential tool for stock market analysts to make informed decisions regarding stock prices. The present research makes use of various exponential smoothing forecasting methods. These include exponential smoothing with multiplicative errors and additive trend (MAN), exponential smoothing with multiplicative errors (MNN), and simple exponential smoothing with additive errors (ANN) for the forecasting of the stock prices of six different companies in the petroleum, electricity, and gas industries that are listed in the IBEX35 index. The database employed for this research contained the IBEX35 index values and stock closing prices from 3 January 2000 to 30 December 2022. The models trained with this data were employed in order to forecast the index value and the closing prices of the stocks under study from 2 January 2023 to 24 March 2023. The results obtained confirmed that although none of the proposed models outperformed the rest for all the companies, it is possible to calculate forecasting models able to predict a 95% confidence interval about real stock closing values and where the index will be in the following three months.

Comparison of Active COVID-19 Cases per Population Using Time-Series Models

This research explored the precision of diverse time-series models for COVID-19 epidemic detection in all the 36 different states and the Federal Capital Territory (FCT) in Nigeria with the maximum count of daily cumulative of confirmed, recovered, and death cases as of 4 November 2020. A 14step forecast system for active coronavirus cases was built, analyzed, and compared for six different deep learning-stimulated and statistical time-series models using two openly accessible datasets. The results obtained showed that based on RMSE metric, ARIMA model obtained the best values for four of the states (0.002537, 0.001969.12E-058, 5.36E-05 values for Lagos, FCT, Edo and Delta states, respectively). While no method is all-encompassing for predicting daily active coronavirus cases for different states in Nigeria, ARIMA model obtains the highest-ranking prediction performance and attained a good position results in other states.

Time Series Analysis Using Different Forecast Methods and Case Fatality Rate for Covid‐19 Pandemic

This study presents forecasting methods using time series analysis for confirmed cases, the number of deaths and recovery cases, and individual vaccination status in different states of India. It aims to forecast the confirmed cases and mortality rate and develop an artificial intelligence method and different statistical methodologies that can help predict the future of Covid‐19 cases. Various forecasting methods in time series analysis such as ARIMA, Holt's trend, naive, simple exponential smoothing, TBATS, and MAPE are extended for the study. It also involved the case fatality rate for the number of deaths and confirmed cases for respective states in India. This study includes the forecast values for the number of positive cases, cured patients, mortality rate, and case fatality rate for Covid‐19 cases. Among all forecast methods involved in this study, the naive and simple exponential smoothing method shows an increased number of positive instances and cured patients.Alternate :Este estudio presenta métodos de pronóstico que utilizan el análisis de series temporales para los casos confirmados, el número de muertes y casos recuperados, y el estado de vacunación individual en diferentes estados de la India. Su objetivo es pronosticar los casos confirmados y la tasa de mortalidad y desarrollar un método de inteligencia artificial y diferentes metodologías estadísticas que puedan ayudar a predecir el futuro de los casos de Covid‐19. Para el estudio se adaptaron varios métodos de pronóstico para el análisis de series temporales como ARIMA, la tendencia de Holt, el ingenuo, el suavizado exponencial simple, TBATS y MAPE. También se incluyó la tasa de fatalidades para el número de muertes y casos confirmados para los respectivos estados de la India. Este estudio incluye los valores de pronóstico para el número de casos positivos, los pacientes curados, la tasa de mortalidad y la tasa de fatalidades para los casos de Covid‐19. Entre todos los métodos de pronóstico utilizados en este estudio, el método ingenuo y el de suavización exponencial simple muestran un mayor número de casos positivos y de pacientes curados.Alternate :抄録本研究は、インドの州における確定症例、死亡数及び回復例、および個人のワクチン接種状況に関する時系列分析を用いた予測方法を提示する。確定症例と死亡率を予測し、人工知能を用いた方法とCOVID‐19の症例の将来を予測するのに役立ついくつかの統計学的方法論を開発することを目指す。ARIMA、Holtのトレンド、単純法、単純指数平滑化法、TBATS、MAPEなどの時系列解析における各種予測法を拡張した。また、インドの各州の死亡者数と確定症例数の致死率も含んだ。本研究は、COVID‐19症例に対する、陽性症例数、治癒患者数、死亡率、および致死率に対する予測値を含む。この研究に含まれるすべての予測法の中で、単純法と単純指数平滑法は、陽性者数と治癒患者数の増加を予測した。

Strategy for Services Anticipation at the Rehabilitation Center of BNN Based on Moving Average and Exponential Smoothing Forecasting Method

One of the Indonesian government's responses to the COVID-19 pandemic is making policies related to restrictions on public services which affects the organizational resilience of the Rehabilitation Center of the National Narcotics Board (BNN). This research aimed to determine the historical pattern of the influence of public service policies during the COVID-19 pandemic on the client population, to forecast the client population for 3 (three) months ahead, and to analyze strategies for anticipating rehabilitation services at the Rehabilitation Center of BNN. This research method is quantitative by using a moving average (MA) and exponential smoothing forecasting model. Based on the validity test, MA is the best forecasting model, which indicates a possibility of a spike in male clients with the same amount in the pre-pandemic period, as many as 310 people, and the average female client is 7 people. Meanwhile, adolescent clients show inaccurate prediction results with MAPE 149.825. Strategies that can be implemented to anticipate a spike in the number of clients if it reaches the highest forecasting point are: increasing the budget, modifying the rehabilitation program for female and adolescent clients, a balanced staff composition, and the availability of facilities and infrastructure. [ FROM AUTHOR] Copyright of Technium Social Sciences Journal is the property of Technium Press Constanta and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full . (Copyright applies to all s.)

Omicron Case Prediction using Machine Learning

It's been over two years that the world has been dealing with the novel Coronavirus Disease 2019 (COVID-19). It has rocked the world in the face of another major outbreak. Countries have undergone various lockdowns curfews in their own ways, which certainly has impacted our daily lives. COVID-19 has undergone various mutations till now. It is responsible for the spikes in COVID-19 cases across the world. The latest variant 'Omicron'., labeled as B.1.1.529, has been marked as a Variant of Concern by the World Health Organization (WHO). It has been proven to be the most infectious, but less deadly as of now. This paper attempts to propose an analysis and prediction of Omicron daily cases in India using SARIMA Exponential Smoothing Machine Learning models. Both of these machine learning models are based on the time series forecasting concept and rely on previous data to predict future outcomes. © 2022 IEEE.

Forecasting Offline Retail Sales in the COVID-19 Pandemic Period: A Case Study of a Complex Shopping Mall in South Korea

This study examines the case of a shopping mall in Seoul, South Korea, based on its offline retail sales data during the period of the enforcement of the COVID-19 pandemic social distancing policy. South Korea implemented strict social distancing, especially in retail categories where people eat out, due to the danger of spreading infectious disease. A total of 55 retail shops' sales data were analyzed and classified into five categories: fashion, food and beverage (f&b), entertainment, cosmetics and sport. Autoregressive integrated moving average (ARIMA) and exponential smoothing (ETS) models were employed, and the autocorrelation (ACF) and partial autocorrelation (PACF) of each retail category's sales data were analyzed. The mean absolute percentage error (MAPE) was used to determine the most suitable forecasting model for each retail category. In this way, the f&b and entertainment retail categories, in which people eat out, were found to have been significantly impacted, with their 2022 sales forecasted to be less than 80% of their 2018 and 2019 sales. The fashion retail category was also significantly impacted, slowly recovering sales in 2022. The cosmetics and sport retail categories were little impacted by the COVID-19 outbreak, with their retail sales having already recovered by 2022. © 2023 by the authors.

The effect of vaccine on COVID-19 spread by function-on-scalar regression model: a case study of Africa.

Aim: This paper aimed to study the effect of the vaccine on the reproduction rate of coronavirus in Africa from January 2021 to November 2021. Subject and methods: Functional data analysis (FDA), a relatively new area in statistics, can describe, analyze, and predict data collected over time, space, or other continuum measures in many countries every day and is increasingly common across scientific domains. For our data, the first step of functional data is smoothing. We used the B-spline method to smooth our data. Then, we apply the function-on-scalar and Bayes function-on-scalar models to fit our data. Results: Our results indicate a statistically significant relationship between the vaccine and the rate of virus reproduction and spread. When the vaccination rate falls, the reproduction rate also decreases. Furthermore, we found that the effect of latitude and the region on the reproduction rate depends on the region. We discovered that in Middle Africa, from the beginning of the year until the end of the summer, the impact is negative, implying that the virus spread due to a decrease in the vaccination rates. Conclusion: The study found that vaccination rates significantly impact the virus's reproduction rate.

The COVID-19 shock and consumer credit: Evidence from credit card data.

We study the dynamic effect of the COVID-19 shock on credit card use in 2020. Local case incidence had a strong negative effect on credit card spending in the early months of the pandemic, which diminished over time. This time-varying pattern was driven by the fear of the virus, rather than government support programs, consistent with the "pandemic fatigue" of consumers. Local pandemic severity also had a strong effect on credit card repayments. These spending and repayment effects offset each other, resulting in no effect on credit card borrowing, consistent with credit-smoothing behavior. The local stringency of nonpharmaceutical interventions also had a negative effect on spending and repayments, albeit smaller in magnitude. We conclude that the pandemic itself was a more important driver of changes in credit card use than the public health policy response.

Raspberry-based Low-resolution Thermal image system using a Smoothing Filter-based Kalman

Since the emergence of global epidemics such as SARS-CoV-2, H1N1, SARS and MERS, a wide range of systems for measuring temperature have been developed based on computer vision to reduce and prevent the virus contagious. By implementing a Raspberry-based Low-resolution embedded system based and a FLIR Lepton® sensor human body temperature is measured and improved by four different algorithms implemented. Firstly, three traditional time-series processes solving such as, Simple Mean (SM), Simple Moving Average (SMA), and Multi Lineal Regression (MLR), and secondly, and online filter-based Kalman predictor were implemented to increase the signal to noise ratio of the acquired temperature magnitude. Results of average prediction for different benchmarks demonstrate the best performance of Kalman Filter upon traditional processes. In addition, this algorithm achieves to smooth output temperature with fewer samples (∼10% of total samples) in comparison MLR and SMA. Finally, Raspberry-based Low-resolution Thermal image system is a feasible tool as a high-speed temperature estimator, by implementation of algorithms codified in Python language. © 2022 IEEE.

Forecasting the number of road accidents in Poland by day of the week and the impact of pandemics and pandemic-induced changes

The COVID19 pandemic has significantly affected the performance of the transport sector and its overall intensity. Reduced mobility has a large impact on the number of road accidents. The aim of this study is to forecast the number of road accidents in Poland and to assess the impact of the COVID19 pandemic on the variation in road crashes. For this purpose, day-wise historical crash data from 2011 onwards have been collected and analysed. Based on real historical field data, the future has been forecasted for both pandemic and nonpandemic variants. Forecasting of the number of accidents has been carried out using selected time series models and exponential models. Based on obtained data, it can be stated that pandemic resulted in a decrease in number of road accidents in Poland with ranges of reduction varying from 11% to 30% based on different days of week. Most visible decrease is observed on 3 days viz. Monday, Wednesday, and Saturday. Further, the projections show that in view of the current situation one may expect further decrease in the number of road accidents in Poland. © 2023 The Author(s). This open access article is distributed under a Creative Commons Attribution (CC-BY) 4.0 license.

Forecasting Combination of Hierarchical Time Series: A Novel Method with an Application to CoVid-19

Multiple, hierarchically organized time series are routinely submitted to the forecaster upon request to provide estimates of their future values, regardless the level occupied in the hierarchy. In this paper, a novel method for the prediction of hierarchically structured time series will be presented. The idea is to enhance the quality of the predictions obtained using a technique of the type forecast reconciliation, by applying this procedure to a set of optimally combined predictions, generated by different statistical models. The goodness of the proposed method will be evaluated using the official time series related to the number of people tested positive to the SARS-CoV-2 in each of the Italian regions, between February 24th 2020 and August 31th 2020. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

THE STOCHASTIC APPROXIMATION METHOD FOR SEMI-RECURSIVE MULTIVARIATE KERNEL-TYPE REGRESSION ESTIMATION

In this research paper, we elaborate an extension of the semi-recursive kernel-type regression function estimator. We investigate the asymptotic properties of this estimator and compare them with non-recursive Nadaraya Watson regression estimator. From this perspective, we first calculate the bias and the variance of the proposed estimator which strongly depend on the choice of three parameters, namely the stepsizes (βn) and (γn) as well as the bandwidth (hn) chosen using one of the best methods of bandwidth selection, the bootstrap approach compared to the plug-in method. An appropriate choice of those parameters yields that, under some conditions, the MSE (Mean Squared Error) of the proposed estimator can be smaller than that of Nadaraya Watson's estimator. We corroborate our theoretical results through simulations studies and by considering two real dataset applications, the French Hospital Data of COVID-19 epidemic as well as the Plasmodium Falciparum Parasite Load (PL). © 2022 Ukrainian National Academy of Sciences. All rights reserved.

Modeling the positive testing rate of COVID-19 in South Africa using a semi-parametric smoother for binomial data.

Identification and isolation of COVID-19 infected persons plays a significant role in the control of COVID-19 pandemic. A country's COVID-19 positive testing rate is useful in understanding and monitoring the disease transmission and spread for the planning of intervention policy. Using publicly available data collected between March 5th, 2020 and May 31st, 2021, we proposed to estimate both the positive testing rate and its daily rate of change in South Africa with a flexible semi-parametric smoothing model for discrete data. There was a gradual increase in the positive testing rate up to a first peak rate in July, 2020, then a decrease before another peak around mid-December 2020 to mid-January 2021. The proposed semi-parametric smoothing model provides a data driven estimates for both the positive testing rate and its change. We provide an online R dashboard that can be used to estimate the positive rate in any country of interest based on publicly available data. We believe this is a useful tool for both researchers and policymakers for planning intervention and understanding the COVID-19 spread.

Specification testing for ordinary differential equation models with fixed design and applications to COVID-19 epidemic models.

Checking the models about the ongoing Coronavirus Disease 2019 (COVID-19) pandemic is an important issue. Some famous ordinary differential equation (ODE) models, such as the SIR and SEIR models have been used to describe and predict the epidemic trend. Still, in many cases, only part of the equations can be observed. A test is suggested to check possibly partially observed ODE models with a fixed design sampling scheme. The asymptotic properties of the test under the null, global and local alternative hypotheses are presented. Two new propositions about U-statistics with varying kernels based on independent but non-identical data are derived as essential tools. Some simulation studies are conducted to examine the performances of the test. Based on the available public data, it is found that the SEIR model, for modeling the data of COVID-19 infective cases in certain periods in Japan and Algeria, respectively, maybe not be appropriate by applying the proposed test.

COVID-19 Future Forecasting Using Machine Learning Models

Covid-19 is a very infectious virus. According to World Health Organization (WHO), millions of individuals have been diagnosed with Covid-19 since then, and at least a million have died as the virus has expanded dramatically. While most of the news on this front is scary, technology is helping to pave the path through this crisis. Manual forecasting is a difficult challenge for humans due to its large scale and complexity. Machine Learning (ML) techniques can effectively predict Covid-19 infected patients. There are a lot of study that have been developed to predict and forecast the future number of cases affected by Covid-19. In this area, our forecasting can be tackled as a problem of supervised learning. Supervised ML is very popular regression methods due to its simplicity to be interpreted by Humans. In this paper, we use two datasets to predict the symptoms through two different types of regression algorithms (single and multiple regression), the ML algorithms are LR, SVM, LASSO, ES and Polynomial regression, for the multiple regression we used LR, SVM and LASSO. The obtained results validate that for the single regression the Exponential Smoothing (ES) outperforms other machine learning approaches like Linear Regression (LR) and LASSO in terms of R-Square, Adjusted R-Square, Mean Absolute Error (MAE), Mean Square Error (MSE), and Root Mean Square Error (RMSE). The same accuracy is observed for the models used in the multiple regression. © 2022 IEEE.