Join our two invited speakers, Mr. Chenwei Bai and Mr. Zhiyong Jin for their talk on Thursday, April 8th.


Time: 2:30pm on Thursday, April 8

Speaker: Mr. Chenwei Bai, Department of Statistics, University of Manitoba

Title of the talk: Comparison of Forecasting Methods for Covid-19

Abstract: Covid-19 is a global challenge that requires research from all areas. However, the main concern for public is when this pandemic will be over. Forecasting the future infectious cases is crucial for answering such question. In this work, we applied Exponential smoothing methods, Autoregressive Integrated Moving Average (ARIMA) Model, and Feed-forward neural network (FNN) to predict the number of confirmed cases and daily increase cases for the next 15 days. Different regions and time ranges were also considered as we tested our models. Root mean square errors (RMSEs) were calculated for each model and used to judge which model performed best. Our results and intervals can provide useful information to the policy makers, and the ultimate goal for us is to find a flexible and robust method for forecasting Coronavirus disease.


Time: 3:00pm on Thursday, April 8

Speaker: Mr. Zhiyong Jin, Department of Statistics, University of Manitoba

Title of the talk: Stochastic Simulation Models for the Spread of Infectious Diseas

Abstract: Developed from the compartment models for the spread of infectious diseases, a stochastic simulation model is raised to simulate the spread of Covid-19 under different patterns of social contact and public health measures in a certain area. We studied several ways of modeling social structure of contacts. A little ball hitting model is used under different assumptions of the moving trace of individuals, such as individuals independently moving uniformly, by Brownian motion or by Levy flight; a network model is used for fixed social relations and fixed social contacts; third-party data such as cellphone location data can also be used under this model frame to better describe the movements of individuals. Compared with the traditional epidemiological compartment models, this simulation model provides a more flexible and realistic approach. It helps to explain the superspreading events, to check the efficacy of public measures and to predict the second wave of the pandemic or even periodic waves.

DATE: Thursday, April 4th, 2021

WHERE: Zoom (see below for more information)

WHEN: 2:30pm – waiting room will open at 2:15pm.

If you wish to attend a Statistics Seminar please contact Po Yang ( from the U of M’s Statistics Department for more details and Zoom meeting information.