A new hybrid ensemble machine-learning model for severity risk assessment and post-COVID prediction system.

Starting from December 2019, the COVID-19 pandemic has globally strained medical resources and caused significant mortality. It is commonly recognized that the severity of SARS-CoV-2 disease depends on both the comorbidity and the state of the patient's immune system, which is reflected in several biomarkers. The development of early diagnosis and disease severity prediction methods can reduce the burden on the health care system and increase the effectiveness of treatment and rehabilitation of patients with severe cases. This study aims to develop and validate an ensemble machine-learning model based on clinical and immunological features for severity risk assessment and post-COVID rehabilitation duration for SARS-CoV-2 patients. The dataset consisting of 35 features and 122 instances was collected from Lviv regional rehabilitation center. The dataset contains age, gender, weight, height, BMI, CAT, 6-minute walking test, pulse, external respiration function, oxygen saturation, and 15 immunological markers used to predict the relationship between disease duration and biomarkers using the machine learning approach. The predictions are assessed through an area under the receiver-operating curve, classification accuracy, precision, recall, and F1 score performance metrics. A new hybrid ensemble feature selection model for a post-COVID prediction system is proposed as an automatic feature cut-off rank identifier. A three-layer high accuracy stacking ensemble classification model for intelligent analysis of short medical datasets is presented. Together with weak predictors, the associative rules allowed improving the classification quality. The proposed ensemble allows using a random forest model as an aggregator for weak repressors' results generalization. The performance of the three-layer stacking ensemble classification model (AUC 0.978; CA 0.920; F1 score 0.921; precision 0.924; recall 0.920) was higher than five machine learning models, viz. tree algorithm with forward pruning; Naïve Bayes classifier; support vector machine with RBF kernel; logistic regression, and a calibrated learner with sigmoid function and decision threshold optimization. Aging-related biomarkers, viz. CD3+, CD4+, CD8+, CD22+ were examined to predict post-COVID rehabilitation duration. The best accuracy was reached in the case of the support vector machine with the linear kernel (MAPE = 0.0787) and random forest classifier (RMSE = 1.822). The proposed three-layer stacking ensemble classification model predicted SARS-CoV-2 disease severity based on the cytokines and physiological biomarkers. The results point out that changes in studied biomarkers associated with the severity of the disease can be used to monitor the severity and forecast the rehabilitation duration.

Modelling and predicting the spread of COVID-19 cases depending on restriction policy based on mined recommendation rules.

This paper is an extended and supplemented version of the paper "Recommendation Rules Mining for Reducing the Spread of COVID-19 Cases", presented by the authors at the 3rd International Conference on Informatics & Data-Driven Medicine in November 2020. The paper examines the impact of government restrictive measures on the spread and effects of COVID-19. The work is devoted to the improvement of recommendation rules based on novel ensemble of machine learning methods such as regression tree and clustering. The dynamics of migration between countries in clusters, and their relationship with the number of confirmed cases and the percentage of deaths caused by COVID-19, were studied on the example of Poland, Italy and Germany. It is shown that there is a clear relationship between the cluster number and the number of new cases of diseases and death. It has also been shown that different countries' policies to prevent the disease, in particular the timing of restrictive measures, correlate with the dynamics of the spread of COVID-19 and the consequences of the disease. For example, the results show a clear proactive tactic of restrictive measures by example of Germany, and catching up on the spread of the disease by example of Italy. A regression tree and guidelines about influence of features on the spreading of COVID-19 and mortality due to this infection have been constructed. The paper predicts the number of deaths due to COVID-19 on a 21-day interval using the obtained guidelines on the example of Sweden. Such forecasting was carried out for two potential government action options: with existing precautionary actions and the same precautionary actions, if they had been taken 20 days earlier (following the example of Germany). The RMSE of the mortality forecast does not exceed 4.2, which shows a good prognostic ability of the developed model. At the same time, the simulation based on the strategy of anticipatory introduction of restrictions gives 2-6% lower values of the forecast of the number of new cases. Thus, the results of this study provide an opportunity to assess the impact of decisions about restrictive measures and predict, simulate the consequences of restrictions policy.