[Paper Review] Predictions of 2019-nCoV Transmission Ending via Comprehensive Methods

Tianyu Zeng, Yunong Zhang+3 more|arXiv (Cornell University)|Feb 12, 2020

COVID-19 diagnosis using AI16 references36 citations

TL;DR

The paper proposes a multi-model framework (MMODEs-NN) and model-free methods to predict interprovincial COVID-19 transmission in China, suggesting deceleration before Feb 18 and ending before April 2020.

ABSTRACT

Since the SARS outbreak in 2003, a lot of predictive epidemiological models have been proposed. At the end of 2019, a novel coronavirus, termed as 2019-nCoV, has broken out and is propagating in China and the world. Here we propose a multi-model ordinary differential equation set neural network (MMODEs-NN) and model-free methods to predict the interprovincial transmissions in mainland China, especially those from Hubei Province. Compared with the previously proposed epidemiological models, the proposed network can simulate the transportations with the ODEs activation method, while the model-free methods based on the sigmoid function, Gaussian function, and Poisson distribution are linear and fast to generate reasonable predictions. According to the numerical experiments and the realities, the special policies for controlling the disease are successful in some provinces, and the transmission of the epidemic, whose outbreak time is close to the beginning of China Spring Festival travel rush, is more likely to decelerate before February 18 and to end before April 2020. The proposed mathematical and artificial intelligence methods can give consistent and reasonable predictions of the 2019-nCoV ending. We anticipate our work to be a starting point for comprehensive prediction researches of the 2019-nCoV.

Motivation & Objective

• Motivate the need for comprehensive prediction approaches beyond traditional epidemiological models.
• Develop a multi-model ODE-based neural network (MMODEs-NN) to simulate interprovincial transmission.
• Introduce fast, linear model-free methods using sigmoid, Gaussian, and Poisson distributions.
• Assess the impact of control policies and travel patterns on transmission dynamics and end timing.

Proposed method

• Introduce MMODEs-NN to simulate transportations via an ODE activation method.
• Apply model-free methods based on sigmoid, Gaussian, and Poisson distributions for rapid predictions.
• Ensure methods are linear and computationally efficient for reasonable forecasting.
• Conduct numerical experiments to evaluate predictions in the context of early 2020 outbreaks.
• Analyze how special policies influenced provincial transmission dynamics.

Experimental results

Research questions

• RQ1 Can interprovincial transmission in mainland China be accurately predicted using MMODEs-NN and model-free approaches?
• RQ2 Do control policies and the Spring Festival travel rush affect the timing of transmission deceleration and ending?
• RQ3 What end time is predicted for the 2019-nCoV outbreak under these comprehensive methods?
• RQ4 How do these methods compare with traditional epidemiological models in forecasting accuracy?

Key findings

• Special policies in some provinces are successful in controlling transmission.
• The transmission is predicted to decelerate before February 18, 2020.
• The transmission is predicted to end before April 2020.
• The proposed methods provide consistent and reasonable predictions for the 2019-nCoV ending.