An ANN-based design for weather derivatives in consideration of meteorological uncertainties

摘要

In this paper, an intelligent design method of weather derivatives is proposed to deal with meteorological uncertainties. This paper aims at equalizing the payoffs in the color option. The proposed method makes use of a hybrid intelligent system that consists of DA (Deterministic Annealing) clustering as a prefiltering technique, MLP (Multi-Layer Perceptron) of ANN (Artificial Neural Network) for a predictor and EPSO (Evolutionary Particle Swarm Optimization) of evolutionary computation for optimizing parameters of the contract. The objective function is to minimize the difference of received payoffs between two companies. The effectiveness of the proposed method is demonstrated in real data simulation.