Improved optimisation of density forecasting models using multi-objective search with applications in risk management

Abstract

Density forecasting is becoming an increasingly popular method of prediction. Density forcasting models produce a probability density function estimate of a future event rather than a standard point estimate. Given that predictions are inherently uncertain, density forecasting is a logical means of presenting predictions and is the preferred forecasting method in most real-world domains where prediction models are widely used e.g. meteorology and finance. The additional information provided by a density forecast enables the prediction user to make optimal decisions under uncertainty.