Continuous traveling time prediction using Genetic Network Programming-based data mining and Neural Networks
Qin Zhang, Huiyu Zhou, Shingo Mabu, Kotaro Hirasawa · Society of Instrument and Control Engineers of Japan · 2012
In this paper, a method combining Genetic Network Programming-based class association rule mining and Neural Networks is proposed for continuous traveling time prediction. Genetic Network Programming (GNP)[1], as an extended algorithm of GP[2], shows its advantage because of its graph structures. GNP is used to generate class association rules[3]. Then, the average matching degree of the data with the rules is calculated. Lastly, the back propagation algorithm of Neural Networks[4] is utilized in order to acquire the concrete prediction of the traveling time.