Time series prediction for the epidemic trends of COVID-19 using the improved LSTM deep learning method: Case studies in Russia, Peru and Iran

Abstract

Highlights A predicting model for the long-term epidemic trend of COVID-19 by using LSTM with rolling update mechanism is proposed. The 150-days ahead epidemic trend of COVID-19 in Russia, Peru and Iran are estimated by our proposed model. The results provide that the epidemic of Peru will end in early December. The number of daily cases in Russia and Iran is expected to fall below 2000 and 1000 by mid-November and early December. By introducing diffusion index, the effectiveness of preventive measures taken by the government are analyzed.

Prediction of COVID-19 confirmed cases combining deep learning methods and Bayesian optimization

“…Besides, the machine and deep learning techniques such as ANN and LSTM have exhibited improvements in COVID-19 time series forecasting studies (e.g. [ 2 , 10 , 12 ]). Also, some methods based on fuzzy logic have been proposed in the literature(e.g.…”

“…Also, in [9] , a new ensemble approach based on ANNs and fuzzy aggregation was proposed and its performance was evaluated on COVID-19 time series of Mexico and its 12 states which showed significant improvement than single ANN. In recent studies [ 2 , [10] , [11] , [12] ], deep learning methods such as LSTM and bidirectional LSTM (BiLSTM) have been utilized for COVID-19 time series forecasting . The results indicated that LSTM and its variants have good performance in predicting the COVID-19 time series.…”

“…Most of those studies (e.g. [ 2 , 10 , 12 ]) have implemented models using hand-tuned hyperparameters. As another contribution, in this study, we utilize the Bayesian Optimization method [17] in order to optimize the hyperparameters of Multi-Head Attention, LSTM, and CNN.…”

What Can COVID-19 Teach Us about Using AI in Pandemics?

“…Need for prompt human validation and response [1,[9][10][11][12]15,17,26,[63][64][65][66][67] Can AI mitigate the spread of biological diseases and guide early treatment?…”

“…Refined analysis of poorly refined, incomplete, or biased data Abdication of human responsibility for triage decision-making [13][14][15][16][17][18]20,25,26,37,[69][70][71][72] How might AI accelerate development of medical therapies and treatment protocols?…”