The range of predictions for calibrated agent-based simulation models

Agent-based simulation is increasingly used to study systems in many areas of business and science. Using agent-based simulation for prediction could be very valuable. However, these models usually have a lot of parameters which are difficult to measure directly leading to uncertainty as to the best values to use. Obtaining the values for the parameters may require calibration of the model against observed historical output data. This type of problem is an inverse problem and there may be many sets of feasible parameter values giving a wide range of predictions. The work described here investigated the extent of this problem for a word of mouth consumer model.