CausalGNN: Causal-Based Graph Neural Networks for Spatio-Temporal Epidemic Forecasting DOI Open Access
Lijing Wang, Aniruddha Adiga, Jiangzhuo Chen

et al.

Proceedings of the AAAI Conference on Artificial Intelligence, Journal Year: 2022, Volume and Issue: 36(11), P. 12191 - 12199

Published: June 28, 2022

Infectious disease forecasting has been a key focus in the recent past owing to COVID-19 pandemic and proved be an important tool controlling pandemic. With advent of reliable spatiotemporal data, graph neural network models have able successfully model inter-relation between cross-region signals produce quality forecasts, but like most deep-learning they do not explicitly incorporate underlying causal mechanisms. In this work, we employ mechanistic guide learning embeddings propose novel framework -- Causal-based Graph Neural Network (CausalGNN) that learns embedding latent space where input features epidemiological context are combined via mutually mechanism using graph-based non-linear transformations. We design attention-based dynamic GNN module capture spatial temporal dynamics. A is added provide for node ordinary differential equations. Extensive experiments on daily new cases at global, US state, county levels show proposed method outperforms broad range baselines. The learned which incorporates organizes efficient way by keeping parameter size small leading robust accurate performance across various datasets.

Language: Английский

Time series forecasting of Covid-19 using deep learning models: India-USA comparative case study DOI Open Access
Sourabh Shastri, Kuljeet Singh, Sachin Kumar

et al.

Chaos Solitons & Fractals, Journal Year: 2020, Volume and Issue: 140, P. 110227 - 110227

Published: Aug. 20, 2020

Language: Английский

Citations

244
Applications of artificial intelligence in battling against covid-19: A literature review DOI Open Access

Mohammad-H. Tayarani N.

Chaos Solitons & Fractals, Journal Year: 2020, Volume and Issue: 142, P. 110338 - 110338

Published: Oct. 3, 2020

Language: Английский

Citations

196
Prediction of COVID-19 confirmed cases combining deep learning methods and Bayesian optimization DOI Open Access
Hossein Abbasimehr, Reza Paki

Chaos Solitons & Fractals, Journal Year: 2020, Volume and Issue: 142, P. 110511 - 110511

Published: Nov. 28, 2020

Language: Английский

Citations

162
Forecasting of COVID-19 cases using deep learning models: Is it reliable and practically significant? DOI Creative Commons

Jayanthi Devaraj,

Rajvikram Madurai Elavarasan,

Rishi Pugazhendhi

et al.

Results in Physics, Journal Year: 2021, Volume and Issue: 21, P. 103817 - 103817

Published: Jan. 14, 2021

The ongoing outbreak of the COVID-19 pandemic prevails as an ultimatum to global economic growth and henceforth, all society since neither a curing drug nor preventing vaccine is discovered. spread increasing day by day, imposing human lives economy at risk. Due increased enormity number cases, role Artificial Intelligence (AI) imperative in current scenario. AI would be powerful tool fight against this predicting cases advance. Deep learning-based time series techniques are considered predict world-wide advance for short-term medium-term dependencies with adaptive learning. Initially, data pre-processing feature extraction made real world dataset. Subsequently, prediction cumulative confirmed, death recovered modelled Auto-Regressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM), Stacked (SLSTM) Prophet approaches. For long-term forecasting multivariate LSTM models employed. performance metrics computed results subjected comparative analysis identify most reliable model. From results, it evident that algorithm yields higher accuracy error less than 2% compared other algorithms studied metrics. Country-specific city-specific India Chennai, respectively, predicted analyzed detail. Also, statistical hypothesis correlation done on datasets including features like temperature, rainfall, population, total infected area population density during months May, June, July August find out best suitable Further, practical significance elucidated terms assessing characteristics, scenario planning, optimization supporting Sustainable Development Goals (SDGs).

Language: Английский

Citations

155
Review on COVID‐19 diagnosis models based on machine learning and deep learning approaches DOI Open Access
Zaid Abdi Alkareem Alyasseri, Mohammed Azmi Al‐Betar, Iyad Abu Doush

et al.

Expert Systems, Journal Year: 2021, Volume and Issue: 39(3)

Published: July 28, 2021

COVID-19 is the disease evoked by a new breed of coronavirus called severe acute respiratory syndrome 2 (SARS-CoV-2). Recently, has become pandemic infecting more than 152 million people in over 216 countries and territories. The exponential increase number infections rendered traditional diagnosis techniques inefficient. Therefore, many researchers have developed several intelligent techniques, such as deep learning (DL) machine (ML), which can assist healthcare sector providing quick precise diagnosis. this paper provides comprehensive review most recent DL ML for studies are published from December 2019 until April 2021. In general, includes 200 that been carefully selected publishers, IEEE, Springer Elsevier. We classify research tracks into two categories: present public datasets established extracted different countries. measures used to evaluate methods comparatively analysed proper discussion provided. conclusion, diagnosing outbreak prediction, SVM widely mechanism, CNN mechanism. Accuracy, sensitivity, specificity measurements previous studies. Finally, will guide community on upcoming development inspire their works future development. This

Language: Английский

Citations

141
Artificial Intelligence for COVID-19: A Systematic Review DOI Creative Commons

Lian Wang,

Yonggang Zhang, Dongguang Wang

et al.

Frontiers in Medicine, Journal Year: 2021, Volume and Issue: 8

Published: Sept. 30, 2021

Background: Recently, Coronavirus Disease 2019 (COVID-19), caused by severe acute respiratory syndrome virus 2 (SARS-CoV-2), has affected more than 200 countries and lead to enormous losses. This study systematically reviews the application of Artificial Intelligence (AI) techniques in COVID-19, especially for diagnosis, estimation epidemic trends, prognosis, exploration effective safe drugs vaccines; discusses potential limitations. Methods: We report this systematic review following Preferred Reporting Items Systematic Reviews Meta-Analyses (PRISMA) guidelines. searched PubMed, Embase Cochrane Library from inception 19 September 2020 published studies AI applications COVID-19. used PROBAST (prediction model risk bias assessment tool) assess quality literature related diagnosis prognosis registered protocol (PROSPERO CRD42020211555). Results: included 78 studies: 46 articles discussed AI-assisted COVID-19 with total accuracy 70.00 99.92%, sensitivity 73.00 100.00%, specificity 25 area under curve 0.732 1.000. Fourteen evaluated based on clinical characteristics at hospital admission, such as clinical, laboratory radiological characteristics, reaching 74.4 95.20%, 72.8 98.00%, 55 96.87% AUC 0.66 0.997 predicting critical Nine models predict peak, infection rate, number infected cases, transmission laws, development trend. Eight explore drugs, primarily through drug repurposing development. Finally, 1 article predicted vaccine targets that have develop vaccines. Conclusions: In review, we shown achieved high performance evaluation, prediction discovery enhance significantly existing medical healthcare system efficiency during pandemic.

Language: Английский

Citations

138
Time series prediction of COVID-19 transmission in America using LSTM and XGBoost algorithms DOI Creative Commons

Junling Luo,

Zhong-Liang Zhang, Yao Fu

et al.

Results in Physics, Journal Year: 2021, Volume and Issue: 27, P. 104462 - 104462

Published: June 22, 2021

In this paper, we establish daily confirmed infected cases prediction models for the time series data of America by applying both long short-term memory (LSTM) and extreme gradient boosting (XGBoost) algorithms, employ four performance parameters as MAE, MSE, RMSE, MAPE to evaluate effect model fitting. LSTM is applied reliably estimate accuracy due long-term attribute diversity COVID-19 epidemic data. Using XGBoost model, conduct a sensitivity analysis determine robustness predictive parameter features. Our results reveal that achieving reduction in contact rate between susceptible individuals isolated uninfected individuals, can effectively reduce number cases. By combining restrictive social distancing tracing, elimination ongoing pandemic possible. predictions are based on real with reasonable assumptions, whereas accurate course heavily depends how when quarantine, isolation precautionary measures enforced.

Language: Английский

Citations

115
Beyond explaining: Opportunities and challenges of XAI-based model improvement DOI Creative Commons
Leander Weber, Sebastian Lapuschkin, Alexander Binder

et al.

Information Fusion, Journal Year: 2022, Volume and Issue: 92, P. 154 - 176

Published: Nov. 23, 2022

Explainable Artificial Intelligence (XAI) is an emerging research field bringing transparency to highly complex and opaque machine learning (ML) models. Despite the development of a multitude methods explain decisions black-box classifiers in recent years, these tools are seldomly used beyond visualization purposes. Only recently, researchers have started employ explanations practice actually improve This paper offers comprehensive overview over techniques that apply XAI practically obtain better ML models, systematically categorizes approaches, comparing their respective strengths weaknesses. We provide theoretical perspective on methods, show empirically through experiments toy realistic settings how can help properties such as model generalization ability or reasoning, among others. further discuss potential caveats drawbacks methods. conclude while improvement based significant beneficial effects even not easily quantifiable properties, need be applied carefully, since success vary depending number factors, dataset used, employed explanation method.

Language: Английский

Citations

78
Healthcare predictive analytics using machine learning and deep learning techniques: a survey DOI Creative Commons
Mohammed Badawy, Nagy Ramadan, Hesham A. Hefny

et al.

Journal of Electrical Systems and Information Technology, Journal Year: 2023, Volume and Issue: 10(1)

Published: Aug. 29, 2023

Abstract Healthcare prediction has been a significant factor in saving lives recent years. In the domain of health care, there is rapid development intelligent systems for analyzing complicated data relationships and transforming them into real information use process. Consequently, artificial intelligence rapidly healthcare industry, thus comes role depending on machine learning deep creation steps that diagnose predict diseases, whether from clinical or based images, provide tremendous support by simulating human perception can even diseases are difficult to detect intelligence. Predictive analytics critical imperative industry. It significantly affect accuracy disease prediction, which may lead patients' case accurate timely prediction; contrary, an incorrect it endanger lives. Therefore, must be accurately predicted estimated. Hence, reliable efficient methods predictive analysis essential. this paper aims present comprehensive survey existing approaches utilized identify inherent obstacles applying these domain.

Language: Английский

Citations

60
Role of Machine Learning Techniques to Tackle the COVID-19 Crisis: Systematic Review DOI Creative Commons
Hafsa Bareen Syeda, Mahanazuddin Syed, Kevin W. Sexton

et al.

JMIR Medical Informatics, Journal Year: 2020, Volume and Issue: 9(1), P. e23811 - e23811

Published: Nov. 15, 2020

SARS-CoV-2, the novel coronavirus responsible for COVID-19, has caused havoc worldwide, with patients presenting a spectrum of complications that have pushed health care experts to explore new technological solutions and treatment plans. Artificial Intelligence (AI)-based technologies played substantial role in solving complex problems, several organizations been swift adopt customize these response challenges posed by COVID-19 pandemic.The objective this study was conduct systematic review literature on AI as comprehensive decisive technology fight crisis fields epidemiology, diagnosis, disease progression.A search PubMed, Web Science, CINAHL databases performed according PRISMA (Preferred Reporting Items Systematic Reviews Meta-Analysis) guidelines identify all potentially relevant studies published made available online between December 1, 2019, June 27, 2020. The syntax built using keywords specific AI.The strategy resulted 419 articles during aforementioned period. Of these, 130 publications were selected further analyses. These classified into 3 themes based applications employed combat crisis: Computational Epidemiology, Early Detection Diagnosis, Disease Progression. studies, 71 (54.6%) focused predicting outbreak, impact containment policies, potential drug discoveries, which under Epidemiology theme. Next, 40 (30.8%) applied techniques detect patients' radiological images or laboratory test results Diagnosis Finally, 19 (14.6%) progression, outcomes (ie, recovery mortality), length hospital stay, number days spent intensive unit Progression theme.In review, we assembled current utilized AI-based methods provide insights different themes. Our findings highlight important variables, data types, resources can assist facilitating clinical translational research.

Language: Английский

Citations

139