PoP: Prediction on Predictions with Ensemble Method


H. Hirose, Y. Koyanagi


IAENG International Journal of Applied Mathematics, Vol.44 Issue 2, pp.103-108, June 2009, Advance Online Version Available: 27 May 2014


In observing the widely spread of patients caused by infectious diseases or the increase of the number of failures of equipments, it is crucial to predict the final number of infected patients or failures at earlier stages. To estimate the number of infected patients, the SIR model, the ordinary differential equation model, statistical truncated model are useful. The predicted value for the final number of patients using data until truncation time T becomes a function (trend) of T. To grasp the prediction trend with truncation time, the L-plot is developed here, which is to plot the predicted final value at the truncation time. We consider the use of the L-plot to predict the final number of patients. For example, we have shown to use the decay function. Applying the multiple methodologies to the same data, we can expect better predicted values. This is called the PoP, the prediction on predictions. As one of the PoP method, we propose to use the ensemble method. By applying these methods to the SARS case, we have found that the ensemble method works well as a PoP method.

Key Words
PoP, Pandemic, SIR model, ordinary differ- ential equation model, statistical truncated model, generalized logistic distribution, ensemble method, L-plot, decay function, restricted RMSE, pandemic, SIR model, ordinary differential equation model, statistical truncated model, ensemble method.



Times Cited in Web of Science:

Cited in Books: