SHORT-TERM PREDICTION OF COMMUNITY CYCLE USAGE BY THE RBF NETWORK

Abstract

The user of community cycle business have been increasing. And now the gap between bike supply and users demand is generated to a specific port. The purpose of this study is to gaze at the increase in efficiency of rearrangement of the bicycle by the staff, to predict the short-term use pattern of community cycle by the RBF network, and to examine the accuracy.
As a result, the prediction accuracy is in the ports that are actively use but to decrease, it was possible to be highly accurate predictions on other ports. In addition, by comparison with other models, RBF network is confirmed to be effective in predicting the usage cycles community.