Analysis on Non-ordinal Traffic to local Airport by using Independent Component Analysis
Shinichi Inoue, Makoto Tsukai, Masashi Kuwano, Makoto Okumura · INFRASTRUCTURE PLANNING REVIEW · 2009
Delays and cancellations of flights often occur in cases of thick fog and heavy rain, and the unexpected delays and cancellations significantly trouble the airport users. In this paper, continuously-observed car traffic data were collected at Hiroshima Airport using portable traffic counters. ICA (Independent Component Analysis), which can decompose time-series into statistically independent series with the mixing coefficients, is applied to the continuously observed traffic series. Based on the statistical tests applied to the mixing coefficients and the decomposed original series, the characteristics of unusual temporal traffic to the airport were clarified.