A decision support model of demand forecast for national immunisation vaccines

This paper presents a demand forecast and optimal allocation model to support the decision making of national vaccine purchase. This research attempts to use ARIMA and Neural Network models respectively to forecast two individual values of 'The yearly total number of inoculations' and 'The annual birthrate' for the next year, and then choose the reasonable and better results generated from one of these two models to conduct a further forecast of 'The yearly total demand volume in doses of vaccines'. According to the experimental results, this research indeed may provide a feasible decision model for yearly vaccine procurement for Taiwan's government immunisation authority and establishes an optimal demand forecast model for a specific vaccine like DTP vaccine.