Comparative analysis of Gated Recurrent Units (GRU), long Short-Term memory (LSTM) cells, autoregressive Integrated moving average (ARIMA), seasonal autoregressive Integrated moving average (SARIMA) for forecasting COVID-19 trends

Alexandria Engineering Journal, Journal Year: 2022, Volume and Issue: 61(10), P. 7585 - 7603

Published: Jan. 6, 2022

Several machine learning and deep models were reported in the literature to forecast COVID-19 but there is no comprehensive report on comparison between statistical models. The present work reports a comparative time-series analysis of techniques (Recurrent Neural Networks with GRU LSTM cells) (ARIMA SARIMA) country-wise cumulative confirmed, recovered, deaths. Gated Recurrent Units (GRU), Long Short-Term Memory (LSTM) cells based (RNN), ARIMA SARIMA trained, tested, optimized trends COVID-19. We deployed python optimize parameters which include (p, d, q) representing autoregressive moving average terms model additional seasonal are denoted by (P, D, Q). Similarly, for RNN models' (number layers, hidden size, rate number epochs) deploying PyTorch framework. best was chosen lowest Mean Square Error (MSE) Root Squared (RMSE) values. For most data countries, learning-based outperformed models, an RMSE values that 40 folds less than But some countries (ARIMA, Further, we emphasize importance various factors such as age, preventive measures healthcare facilities etc. play vital role rapid spread pandemic.

Language: Английский

---

Application of machine learning in the prediction of COVID-19 daily new cases: A scoping review

Heliyon, Journal Year: 2021, Volume and Issue: 7(10), P. e08143 - e08143

Published: Oct. 1, 2021

COVID-19 has produced a global pandemic affecting all over of the world. Prediction rate spread and modeling its course have critical impact on both health system policy makers. Indeed, making depends judgments formed by prediction models to propose new strategies measure efficiency imposed policies. Based nonlinear complex nature this disorder difficulties in estimation virus transmission features using traditional epidemic models, artificial intelligence methods been applied for spread. importance machine deep learning approaches spreading trend, present study, we review studies which used these predict number cases COVID-19. Adaptive neuro-fuzzy inference system, long short-term memory, recurrent neural network multilayer perceptron are among mostly regard. We compared performance several Root means squared error (RMSE), mean absolute (MAE), R

Language: Английский

---

Lumpy Skin Disease Outbreaks in Africa, Europe, and Asia (2005–2022): Multiple Change Point Analysis and Time Series Forecast

Viruses, Journal Year: 2022, Volume and Issue: 14(10), P. 2203 - 2203

Published: Oct. 7, 2022

LSD is an important transboundary disease affecting the cattle industry worldwide. The objectives of this study were to determine trends and significant change points, forecast number outbreak reports in Africa, Europe, Asia. report data (January 2005 January 2022) from World Organization for Animal Health analyzed. We determined statistically points using binary segmentation, auto-regressive moving average (ARIMA) neural network (NNAR) models. Four identified each continent. year between third fourth (2016–2019) African was period with highest mean reports. All outbreaks Europe corresponded massive during 2015–2017. Asia had 2019 after detected point 2018. For next three years (2022–2024), both ARIMA NNAR a rise Africa steady Europe. However, predicts stable Asia, whereas increase 2023–2024. This provides information that contributes better understanding epidemiology LSD.

Language: Английский

---

The Impact of the COVID-19 Pandemic on the Global Web and Video Conferencing SaaS Market

Electronics, Journal Year: 2022, Volume and Issue: 11(16), P. 2633 - 2633

Published: Aug. 22, 2022

The COVID-19 pandemic related government interventions produced rapid decreases in worldwide economic and social activity, with multifaceted consequences. In particular, the disruption of key industries significant lifestyle changes aftermath outbreak led to exponential adoption web video conferencing Software as a Service (SaaS) programs solutions-led market growth. However, magnitude persistence impact on solutions segment remain uninvestigated. Building previous evidence linking population web-search behavior, private consumption, retail sales, this study sources employs Google Trends data an analytical forecasting tool for videoconferencing market. It implements univariate forecast evaluation approach that assesses predictive performance several statistical machine-learning models relative search volume (RSV) two SaaS program leaders, Zoom Teams. ETS is found provide best consumer GT interest both RSV series. A baseline level over first wave subsequently further serves estimate excess February 2020–August 2020 period. Results indicate has created or abnormal global would not have occurred absence pandemic. Other findings persistent stabilized at higher levels than pre-pandemic period although saturation detected.

Language: Английский

---

Forecasting of daily new lumpy skin disease cases in Thailand at different stages of the epidemic using fuzzy logic time series, NNAR, and ARIMA methods

Preventive Veterinary Medicine, Journal Year: 2023, Volume and Issue: 217, P. 105964 - 105964

Published: June 16, 2023

Language: Английский

---

Time-Series Analysis for the Number of Foot and Mouth Disease Outbreak Episodes in Cattle Farms in Thailand Using Data from 2010–2020

Viruses, Journal Year: 2022, Volume and Issue: 14(7), P. 1367 - 1367

Published: June 23, 2022

Thailand is one of the countries where foot and mouth disease outbreaks have resulted in considerable economic losses. Forecasting an important warning technique that can allow authorities to establish FMD surveillance control program. This study aimed model forecast monthly number outbreak episodes (n-FMD episodes) using time-series methods, including seasonal autoregressive integrated moving average (SARIMA), error trend seasonality (ETS), neural network autoregression (NNAR), Trigonometric Exponential smoothing state–space with Box–Cox transformation, ARMA errors, Trend Seasonal components (TBATS), hybrid methods. These methods were applied n-FMD (n = 1209) from January 2010 December 2020. Results showed had a stable 2020, but they appeared increase 2014 The followed pattern, predominant peak occurring September November annually. single-technique yielded best-fitting models, SARIMA(1,0,1)(0,1,1)12, NNAR(3,1,2)12,ETS(A,N,A), TBATS(1,{0,0},0.8,{<12,5>}. Moreover, SARIMA-NNAR NNAR-TBATS models performed best on validation datasets. incorporate non-linear better than others. forecasts highlighted rising Thailand, which shares borders several endemic cross-border trading cattle found common. Thus, strategies effective measures prevent should be strengthened not only also neighboring countries.

Language: Английский

---

Analyzing and forecasting poultry meat production and export volumes in Thailand: a time series approach

Cogent Food & Agriculture, Journal Year: 2024, Volume and Issue: 10(1)

Published: July 16, 2024

Amidst global food security challenges driven by population growth and economic fluctuations, the accurate prediction of production has become increasingly important. Given Thailand's position among world's top 10 poultry meat producers exporters, forecasting these figures is essential for effective planning. This study aims to analyze trends seasonal patterns forecast export volumes using various time series models. The data, which included in Thailand its volume from 2017 2023, was analyzed models including SARIMA, NNAR, ETS, TBATS, STL THETA. Forecast were constructed this study, their predictive performances evaluated compared across different results reveal consistent upward volumes. These are complemented patterns, with peaking March exhibiting a similar trajectory. High periods observed annually between September November. In terms accuracy, SARIMA model outperformed other volume, while THETA excels predicting volume. applied volumes, highlighting practical application significance context, thereby providing information planning relevant authorities stakeholders.

Language: Английский

---

Detections and SIR simulations of the COVID-19 pandemic waves in Ukraine

Computational and Mathematical Biophysics, Journal Year: 2021, Volume and Issue: 9(1), P. 46 - 65

Published: Jan. 1, 2021

Abstract Background. Unfortunately, the COVID-19 pandemic is still far from stabilizing. Of particular concern sharp increase in number of diseases June-July, September-October 2020 and February-March 2021. The causes consequences this cases are waiting for their researchers, but there already an urgent need to assess possible duration pandemic, expected patients deaths. Correct simulation infectious disease dynamics needs complicated mathematical models many efforts unknown parameters identification. Constant changes conditions (in particular, peculiarities quarantine its violation, situations with testing isolation patients) cause various epidemic waves, lead parameter values models. Objective. In article, waves Ukraine will be detected, calculated discussed. estimations durations final sizes presented. Methods. We propose a simple method detection based on differentiation smoothed cases. use generalized SIR (susceptible-infected-removed) model waves. known exact solution differential equations statistical approach were used. different data sets accumulated order compare results simulations predictions. Results. Nine detected corresponding optimal identified. spreading infection versus time calculated. probably began January 2020. If current trends continue, end should no earlier than summer Conclusions. cases, identification helpful select make some reliable obtained information useful regulate activities, predict medical economic pandemic.

Language: Английский

---

Forecasting and classification of new cases of COVID 19 before vaccination using decision trees and Gaussian mixture model

Alexandria Engineering Journal, Journal Year: 2022, Volume and Issue: 62, P. 327 - 333

Published: July 8, 2022

Regarding the pandemic taking place in world from spread of Coronavirus and viral mutations, need has arisen to analyze epidemic data terms numbers infected deaths, different geographical regions, dynamics virus. In China, total number reported infections is 224,659 on June 11, 2022. this paper, Gaussian Mixture Model decision tree method were used classify predict new cases Although we focus mainly Chinese case, model general adapted any context without loss validity qualitative results. The Chi-Squared (χ2) Automatic Interaction Detection (CHAID) was applied creating structure, been classified into five classes, according BIC criterion. best mixture E (Equal variance) with components. considered sets health organization (WHO) January 5, 2020, 12, November 2021. We provide numerical results based case.

Language: Английский

---

Endemic–epidemic models to understand COVID-19 spatio-temporal evolution

Spatial Statistics, Journal Year: 2021, Volume and Issue: 49, P. 100528 - 100528

Published: July 12, 2021

Language: Английский

---