A fighter fly out trajectory time series data mining demo, I used convolution operator metric and AGNES algorithm to clustering the flyout data samples into left, straight and right categories. This app includes a fly out trajectory generation program which was developed by matlab and simulink.
1.To dude Sun Yang:
'mts-air-maneuvering-samples-1000-txt-form.txt' time series data description. we have 1000 time series, each serie have 0-9 sample points by every second, so 1000*10=10000, then the data set have 10000 rows. each record in 1 second was composed as below:
x | y | z | y_global_load_factor | time_in_seconds
You could generate more time series samples by using my matlab program 'run.m' and simulink model 'maneuver_generator.slx'
2.Clustering Algorithm adopted by haiyin piao
2.1 time series dictance metric by convolution operator.
2.2 AGNES algorithm.
3.Clustering Results seems good.
Flight Trajectory Hierarchical Clustering Dendrogram