Research on Similarity Mining for Flight Data

Similarity search for time series has attracted much research attention recently. Flight Data is a set of data arranged by chronological order. Through depth study on the characteristics of Flight Data, in this paper a subsequence matching algorithm based on the Discrete Fourier Transform (DFT) for Flight Data is proposed. In order to reduce the number of dimensions for Flight Data, this algorithm uses the Discrete Fourier Transform (DFT) to map the original Flight Data to the frequency domain. Then, the spatial data index structure is built using the adaptive segmentation, and the Euclidean distance is used as the similarity measurement. The experiment results demonstrate that this algorithm has much higher time efficiency and accuracy rate than the sliding window regression algorithm.