[, 3/18/2020]  The management of coronavirus outbreaks (COVID-19) is not only the responsibility of the medical science. Mathematics also has a role in tackling this pandemic outbreak through mathematical modeling.

Professor of the Faculty of Mathematics and Natural Sciences, Universitas Padjadjaran, Prof. Dr. Budi Nurani Ruchjana, M.S., together with a team of researchers from the Stochastic Modeling Expertise Group of the Mathematics Department of the Faculty of Mathematics and Natural Sciences Universitas Padjadjaran have tried to identify the probabilities of the spread of coronavirus using a stochastic model.

“The stochastic model is a model related to probability. We see that everything in nature is random, for example the coronavirus, when it comes to the world is also random, never knowing who will be infected” Prof. Budi said.

By applying spatio-temporal modeling or random observations based on location and time, Prof. Budi simply looks for the probabilities of the spread of coronavirus based on data available on the page. The data taken is data infected with coronavirus around the world in the span of 23 January to 9 March 2020.

The first data analysis was performed to determine the probability of a situation based on the previous situation from the observation of the number of coronavirus sufferers observed every day with the condition assumed to be constant and homogeneous in all parts of the globe, as well as if the spatial condition was in the form of the number of sufferers above average (many) and below average (small).

The first data analysis uses the Markov chain stationary distribution, found that in average 2.433 people worldwide affected by the virus per day. The average number is still assumed that the phenomenon of coronavirus outbreaks in each country is the same.

The average data is then calculated by using a Markov chain stationary distribution, with the condition of less than the average is assumed to be small, whereas above average is assumed to be many. Then the initial results were obtained that under average coronavirus sufferers were 53%, whereas sufferers above average were 47%.

“This is still a preliminary study, continuous modeling must be done,” she said.

Furthermore, the calculation is done to determine the prediction of the number of patients based on the location that has not been sampled. This is due to a number of countries, especially those close to China, there is no information infected with the virus. The two countries sampled in this identification are Laos and Myanmar, two countries close to China.

Prof. Budi explained, assuming that the same phenomenon of the spread of coronavirus in the world, there is a possibility that regions close to China are also vulnerable to infection.

The process of finding predictions at these non-sampled locations uses the Ordinary Point Kriging (OK) method. As a result, it is predicted that on average of 3-4 people will be infected with coronavirus in Laos or Myanmar.

“In recommendation, we dare to convey even though it is still on paper in numbers, but this can be a warning to be more vigilant, ” Prof. Budi said.

Prof. Budi admitted that the results of these two calculations still require further analysis. This is due to the data used are assumed to be homogeneous. In addition, multidisciplinary research collaboration is also needed.

Nevertheless, she continued, this preliminary data could be a picture to increase awareness of the coronavirus pandemic. Predictions in non-sampled locations are not intended to trigger panic, but to increase future vigilance.

The process of calculating opportunities and predictions in non-sampled locations is part of the Workshop “Spatiotemporal Modeling for Prediction of Coronavirus Patients (COVID-19) in Non-Sampled Locations using R” which was held by KBK in the Stochastic Modeling of the Mathematics Department of the Faculty of Mathematics and Natural Sciences Universitas Padjadjaran in Jatinangor, Friday, March 13, 2020, attended by around 130 participants from West Java and Jakarta.

Prof. Budi said that the workshop was held to commemorate International Mathematics Day on March 14. In addition, this activity is also a realization of the International Consortium of Research Innovation and Staff Exchange_Social Media Analytics (RISE_SMA) in collaboration between researchers of the Department of Mathematics and the Department of Computer Science of the Faculty of Mathematics and Natural Sciences, Universitas Padjadjaran, and the University of Duisburg Essen-Germany (Coordinator) and other tertiary institutions from Leiden, Norway, Brazil, and Australia fully funded by European Union in 2019-2022. (dkd/dfa)*

Share this: