Math Models for Humans vs. Zombies on USU campus
Class
Article
Department
Mathematics and Statistics
Faculty Mentor
Jim Powell
Presentation Type
Poster Presentation
Abstract
In the fall of 2011 Utah State University's campus was invaded by a group of 15 zombies. While the invasion only lasted one week, the "life-impaired" continued to grow in numbers through "aggressive conversion" techniques among the living. At our own peril, we managed to monitor the rate of infection, develop mathematical models and in the end, predict human fate. Disease dynamics explain how an infection spreads throughout a population. The zombie infection, in this case, was incurable. Furthermore, the zombie and human populations are inversely proportional. In other words, as the number or zombies go up, the number of humans go down. From these basic facts, models can be fashioned to predict the amount of zombies and humans at any given point in time. Data was collected from three different semesters: fall 2011, spring 2012 and fall 2012. The models were fit to all three sets of data. A few different kinds of models were used including a logistic model, a theta logistic model, a competition model that accounted for limited resources, and a model that included the delay in reporting of the conversion of humans to zombies. We compared the models using AIC and BIC values and used negative log likelihood to pick the best model. We found that a model that included a competition dependent factor, a delayed reporting effect and a death rate for the zombies gave us the best results. These findings will help us understand and predict how this sort of a disease would spread on USU campus.
Start Date
4-9-2015 1:30 PM
Math Models for Humans vs. Zombies on USU campus
In the fall of 2011 Utah State University's campus was invaded by a group of 15 zombies. While the invasion only lasted one week, the "life-impaired" continued to grow in numbers through "aggressive conversion" techniques among the living. At our own peril, we managed to monitor the rate of infection, develop mathematical models and in the end, predict human fate. Disease dynamics explain how an infection spreads throughout a population. The zombie infection, in this case, was incurable. Furthermore, the zombie and human populations are inversely proportional. In other words, as the number or zombies go up, the number of humans go down. From these basic facts, models can be fashioned to predict the amount of zombies and humans at any given point in time. Data was collected from three different semesters: fall 2011, spring 2012 and fall 2012. The models were fit to all three sets of data. A few different kinds of models were used including a logistic model, a theta logistic model, a competition model that accounted for limited resources, and a model that included the delay in reporting of the conversion of humans to zombies. We compared the models using AIC and BIC values and used negative log likelihood to pick the best model. We found that a model that included a competition dependent factor, a delayed reporting effect and a death rate for the zombies gave us the best results. These findings will help us understand and predict how this sort of a disease would spread on USU campus.