Smallpox Epidemics Using a Computational Model利用一个计算机模型预防天花课件.ppt

Preventing Smallpox Epidemics Using a Computational Model By Chintan Hossain and Hiren Patel Facts About Smallpox Symptoms occur in stages Highly contagious (causes epidemics) Fatal in ?30% cases There is a vaccine - Death may occur GOAL (Objective) Prevent smallpox epidemics via. vaccination. Vaccinate as few as possible because: 1. Minimize reactions 2. Reduce cost Stages of Smallpox Normal (Susceptible) Immune (or vaccinated) Incubation First Stage Early Symptoms Late Symptoms Death Our Model: Social Networks Cliques Represent: Families Workplaces School Our Society Generator Algorithm Use random numbers to pick a family size. Generate a clique of that size. Repeat to create more families. Use a similar technique to generate schools and workplaces. Schools and workplaces connect existing vertices, not new vertices. Our Model Comes Alive! MARKOV GRAPH + SOCIETY NETWORK SIMULATION Advance time 1 day Spread Disease Advance Stages Death Procedure Run the society generator Vaccinate k% of people with most friends (vertices with the greatest degree) Control: k = 0% Variable: Vary percent, k, vaccinated Randomly, infect one person. Run simulation, and observe results (percent infect and length of epidemic) OUR PROGRAM Results Results (cont…) Vaccinating more people decreases the % infected The % infected becomes small if over 50% are vaccinated. Conclusion Vaccinating 50% of the population effectively prevents epidemics. Vaccinating less than 50% may not prevent an epidemic, but it reduces the severity and speed of the epidemic. * * HYPOTHESIS: Vaccinating certain percentage of the population may be sufficient to prevent a smallpox epidemic. Vaccination Contraction 14 days 3 days 9 days Recovery 9 days 0.1% chance / day 0.5% chance / day 3.0% chance / day Normal (Susceptible) Incubation Death First Stage Late Symptoms Early Symptoms Immune Vaccinated \ Normal (Susceptible) Infected Stage Vaccinated / Immune Death ? ? ? FIRST Spread EARLY LATE