Deep learning modeling of public's sentiments towards temporal evolution of COVID-19 transmission.

Public sentiments towards global pandemics are important for public health assessment and disease control. This study develops a modularized deep learning framework to quantify public sentiments towards COVID-19, followed by leveraging the predicted sentiments to model and forecast the daily growth rate of confirmed COVID-19 cases globally, via a proposed G parameter. In the proposed framework, public sentiments are first modeled via a valence dimensional indicator, instead of discrete schemas, and are classified into 4 primary emotional categories: (a) neutral; (b) negative; (c) positive; (d) ambivalent, by using multiple word embedding models and classifiers for text sentiments analyses and classification. The trained model is subsequently applied to analyze large volumes (millions in quantity) of daily Tweets pertaining to COVID-19, ranging from 22 Jan 2020 to 10 May 2020. The results demonstrate that the global community gradually evokes both positive and negative sentiments towards COVID-19 over time compared to the dominant neural emotion at its inception. The predicted time-series sentiments are then leveraged to train a deep neural network (DNN) to model and forecast the G parameter by achieving the lowest possible mean absolute percentage error (MAPE) score of around 17.0% during the model's testing step with the optimal model configuration.

Data-driven multiscale modelling and analysis of COVID-19 spatiotemporal evolution using explainable AI.

To quantificationally identify the optimal control measures for regulators to best minimize COVID-19's growth (G-rate) and death (D-rate) rates in today's context, this paper develops a top-down multiscale engineering approach which encompasses a series of systematic analyses, namely: (global scale) predictive modelling of G-rate and D-rate due to COVID-19 globally, followed by determining the most effective control factors which can best minimize both parameters over time via explainable Artificial Intelligence (AI) with SHAP (SHapley Additive exPlanations) method; (continental scale) same predictive forecasting of G-rate and D-rate in all continents, followed by performing explainable SHAP analysis to determine the most effective control factors for the respective continents; and (country scale) clustering the different countries (> 150 in total) into 3 main clusters to identify the universal set of effective control measures. By using the historical period between 2 May 2020 and 1 Oct 2021, the average MAPE scores for forecasting G-rate and D-rate are within 10%, or less on average, at the global and continental scales. Systematically, we have quantificationally demonstrated that the top 3 most effective control measures for regulators to best minimize G-rate universally are COVID-CONTACT-TRACING, PUBLIC-GATHERING-RULES, and COVID-STRINGENCY-INDEX, while the control factors relating to D-rate depend on the modelling scenario.

A Prediction Model for High Risk of Positive RT-PCR Test Results in COVID-19 Patients Discharged From Wuhan Leishenshan Hospital, China.

Several recent studies have reported that a few patients had positive SARS-CoV-2 RNA tests after hospital discharge. The high-risk factors associated with these patients remain to be identified. A total of 463 patients with COVID-19 discharged from Leishenshan Hospital in Wuhan, China, between February 8 and March 8, 2020 were initially enrolled, and 351 patients with at least 2 weeks of follow-up were finally included. Seventeen of the 351 discharged patients had positive tests for SARS-CoV-2 RNA. Based on clinical characteristics and mathematical modeling, patients with shorter hospital stays and less oxygen desaturation were at higher risk of SARS-CoV-2 RNA reoccurrence after discharge. Notably, traditional Chinese medicine treatment offered extensive benefits to reduce risk. Particular attention should be paid to those patients with high risk, and traditional Chinese medicine should be advocated.

Discovering spatial-temporal patterns via complex networks in investigating COVID-19 pandemic in the United States.

A novel approach combining time series analysis and complex network theory is proposed to deeply explore characteristics of the COVID-19 pandemic in some parts of the United States (US). It merges as a new way to provide a systematic view and complementary information of COVID-19 progression in the US, enabling evidence-based responses towards pandemic intervention and prevention. To begin with, the Principal Component Analysis (PCA) varimax is adopted to fuse observed time-series data about the pandemic evolution in each state across the US. Then, relationships between the pandemic progress of two individual states are measured by different synchrony metrics, which can then be mapped into networks under unique topological characteristics. Lastly, the hidden knowledge in the established networks can be revealed from different perspectives by network structure measurement, community detection, and online random forest, which helps to inform data-driven decisions for battling the pandemic. It has been found that states gathered in the same community by diffusion entropy reducer (DER) are prone to be geographically close and share a similar pattern and tendency of COVID-19 evolution. Social factors regarding the political party, Gross Domestic Product (GDP), and population density are possible to be significantly associated with the two detected communities within a constructed network. Moreover, the cluster-specific predictor based on online random forest and sliding window is proven useful in dynamically capturing and predicting the epidemiological trends for each community, which can reach the highest.

Discovering optimal strategies for mitigating COVID-19 spread using machine learning: Experience from Asia.

To inform data-driven decisions in fighting the global pandemic caused by COVID-19, this research develops a spatiotemporal analysis framework under the combination of an ensemble model (random forest regression) and a multi-objective optimization algorithm (NSGA-II). It has been verified for four Asian countries, including Japan, South Korea, Pakistan, and Nepal. Accordingly, we can gain some valuable experience to better understand the disease evolution, forecast the prevalence of the disease, which can provide sustainable evidence to guide further intervention and management. Random forest with a proper rolling time-window can learn the combined effects of environmental and social factors to accurately predict the daily growth of confirmed cases and daily death rate on a national scale, which is followed by NSGA-II to find a range of Pareto optimal solutions for ensuring the minimization of the infection rate and mortality at the same time. Experimental results demonstrate that the predictive model can alert the local government in advance, allowing the accused time to put forward relevant measures. The temperature in the category of environment and the stringency index belonging to the social factor are identified as the top 2 important features to exert a greater impact on the virus transmission. Moreover, optimal solutions provide references to design the best control strategies towards pandemic containment and prevention that can accommodate the country-specific circumstance, which are possible to decrease the two objectives by more than 95%. In particular, appropriate adjustment of social-related features needs to take priority over others, since it can bring about at least 1.47% average improvement of two objectives compared to environmental factors.

Hybrid deep learning of social media big data for predicting the evolution of COVID-19 transmission.

In this study, a hybrid deep-learning model termed as ODANN, built upon neural networks (NN) coupled with data assimilation and natural language processing (NLP) features extraction methods, has been constructed to concurrently process daily COVID-19 time-series records and large volumes of COVID-19 related Twitter data, as representative of the global community's aggregated emotional responses towards the current pandemic, to model the growth rate in the number of confirmed COVID-19 cases globally via a proposed G parameter. Overall, there were 3 key components to ODANN's development phase, namely: (i) data hydration and pre-processing were performed on COVID-19 related Twitter data ranging between 23 January 2020 and 10 May 2020, which amounted to over 100 million Tweets written in English language; (ii) multiple NLP features extraction methods were subsequently leveraged to encode the hydrated Twitter data into useful semantic word vectors for training ODANN under an optimal set of hyperparameters; and (iii) historical time-series data of defined characteristics were also assimilated into ODANN's selected hidden layer(s) to model the G parameter daily with a lead-time of 1 day. By far, our experimental results demonstrated that by adopting a rolling time-window size of 5 days, with respect to the number of historical time-series records for assimilating different data features, enabled ODANN to outperform other traditional time-series models and recent studies, in terms of the computed RMSE and MAE scores attained from the model's testing step. Overall, the summarized results from ODANN demonstrated its competitive edge in modelling and forecasting the growth rate in the number of COVID-19 cases globally.

Correlating dynamic climate conditions and socioeconomic-governmental factors to spatiotemporal spread of COVID-19 via semantic segmentation deep learning analysis.

In this study, we develop a deep learning model to forecast the transmission rate of COVID-19 globally, via a proposed G parameter, as a function of fused data features which encompass selected climate conditions, socioeconomic and restrictive governmental factors. A 2-step optimization process is adopted for the model's data fusion component which systematically performs the following: (Step I) determining the optimal climate feature which can achieve good precision score (> 70%) when predicting the spatial classes distribution of the G parameter on a global scale consisting of 251 countries, followed by (Step II) fusing the optimal climate feature with 11 selected socioeconomic-governmental factors to further improve the model's predictive capability. By far, the obtained results from the model's testing step indicate that land surface temperature day (LSTD) has the strongest correlation with the global G parameter over time by achieving an average precision score of 72%. When coupled with relevant socioeconomic-governmental factors, the model's average precision score improves to 77%. At the local scale analysis for selected countries, our proposed model can provide insights into the relationship between the fused data features and the respective local G parameter by achieving an average accuracy score of 79%.