Tropical cyclone identification and tracking system using integrated neural oscillatory elastic graph matching and hybrid RBF network track mining techniques

We present an automatic and integrated neural network-based tropical cyclone (TC) identification and track mining system. The proposed system consists of two main modules: 1) TC pattern identification system using neural oscillatory elastic graph matching model; and 2) TC track mining system using hybrid radial basis function network with time difference and structural learning algorithm. For system evaluation, 120 TC cases appeared in the period between 1985 and 1998 provided by National Oceanic and Atmospheric Administration are being used. Comparing with the bureau numerical TC prediction model used by Guam and the enhanced model proposed by Jeng et al. (1991), the proposed hybrid RBF has attained an over 30% and 18% improvement in forecast errors.