Accuracy of COVID-19 evolution models for different forecast horizons

Currently, the statistics on COVID-19 for many regions are accumulated for the time span of over than two years, which facilitates the use of data-driven algorithms, such as neural networks, for prediction of the disease's further development. This article provides a comparative analysis of various forecasting models of COVID-19 dynamics. The forecasting is performed for the period from 07/20/2020 to 05/05/2022 using statistical data from the regions of the Russian Federation and the USA. The forecast target is defined as the sum of confirmed cases over the forecast horizon. Models based on the Exponential Smoothing (ES) method and deep learning methods based on Long Short-Term Memory (LSTM) units were considered. The training data set included the data from all regions available in the full data set. The MAPE metric was used for model comparison, the evaluation of the effectiveness of LSTM in the learning process was carried out using cross-validation on the mean squared error (MSE) metric. The comparisons were made with the models from various literature sources, as well as with the baseline model "tomorrow as today" (for which the sum of cases over the forecast horizon is supposed to be equal to the current case number multiplied by the forecast horizon length). It was shown that on small horizons (up to 28 days) the "tomorrow as today” model and ES algorithms show better accuracy than LSTM. In turn, on longer horizons (28 days or more), the preference should be given to the more complex LSTM-based model. © Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0)

On the accuracy of Covid-19 forecasting methods in Russia for two years

The effectiveness of predicting the dynamics of the coronavirus pandemic for Russia as a whole and for Moscow is studied for a two-year period beginning March 2020. The comparison includes well-proven population models and statistic methods along with a new data-driven model based on the LSTM neural network. The latter model is trained on a set of Russian regions simultaneously, and predicts the total number of cases on the 14-day forecast horizon. Prediction accuracy is estimated by the mean absolute percent error (MAPE). The results show that all the considered models, both simple and more complex, have similar efficiency. The lowest error achieved is 18% MAPE for Moscow and 8% MAPE for Russia.

Baseline Accuracies of Forecasting COVID-19 Cases in Russian Regions on a Year in Retrospect Using Basic Statistical and Machine Learning Methods.

The large amount of data accumulated so far on the dynamics of the COVID-19 outbreak has allowed assessing the accuracy of forecasting methods in retrospect. This work compares several basic time series analysis methods, including machine learning methods, for forecasting the number of confirmed cases for some days ahead. Year-long data for all regions of Russia has been used from the Yandex DataLens platform. As a result, accuracy estimates for these basic methods have been obtained for Russian regions and Russia as a whole, in dependence on the forecasting horizon. The best basic models for forecasting for 14 days are exponential smoothing and ARIMA, with an error of 11-19% by the MAPE metric for the latest part of the course of the epidemic. The accuracies obtained can be considered as baselines for more complex prospective models.

The Russian language corpus and a neural network to analyse Internet tweet reports about Covid-19

This work is aimed at creating a tool for filtering messages from Twitter users by the presence of mentions of coronavirus disease in them. For this purpose, a corpus of Russian-language tweets was created, which contains the part of 10 thousand tweets that are manually divided into several classes with different levels of confidence: potentially have covid, have covid now, other cases, and an unmarked part – 2 million tweets on the topic of the pandemic. The paper presents the process of creating a corpus of tweets from the stage of data collection, their preliminary filtering and subsequent annotation according to the presence of disease description. Machine learning methods were compared according to classification task on tweets. It is shown that a model based on the XLM-RoBERTa topology with additional training on corpus of unmarked tweets gives the F1 score of 0.85 on binary classification task ("potentially have covid have covid now" vs "other"). This is 12% higher relative to the simplest model using TF-IDF encoding and SVM classifier and 5% higher relative to the RuDR-BERT-based model. The created toolkit will expand the feature space of models for predicting the spread of coronavirus infection and other pandemics by adding the dynamics of discussion on social networks, which characterizes people’s attitudes towards it. © Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0).

Baseline accuracy of forecasting COVID-19 cases in Moscow region on a year in retrospect using basic statistical and machine learning methods

The large amount of data that has accumulated so far on the dynamics of the COVID-19 outbreak has allowed to assess the accuracy of forecasting methods in retrospect. This work is devoted to comparing a set of basic time series analysis methods for forecasting the number of confirmed cases for 14 days ahead: machine learning methods, exponential smoothing, autoregressive methods, along with variants of SIR and SEIR. On the year-long data for Moscow, the best basic model is showed to be SEIR within which the basic reproduction number R0 is predicted using a regression model, achieving the mean error of 16% by the MAPE metric. The resulting accuracy can be considered a baseline for a more complex prospective model that would be based on the presented approach. © 2021 Institute of Physics Publishing. All rights reserved.