A feature extraction method based on BSO algorithm for flight data

The feature extraction problem for flight data has aroused increasing attention in the practical and the academic aspects. It can reveal the inherent correlation relation among different parameters for the conditional maintenance of the aircraft. However, the high-dimensional and continuous features in the real number field bring challenges to the extraction algorithms for flight data. Brain Storm Optimization (BSO) algorithm can acquire the optimal solutions by continuously converging and diverging the solution set. In this chapter, a feature extraction method based on BSO algorithm is proposed to mine the associate rules from flight data. By using the designed real-number encoding strategy, the intervals and rule template can be handled directly without data discretization and rule template preset processes. Meanwhile, as the frequent item generation process is unnecessary in our proposed algorithm, the time and space complexity will be reduced simultaneously. In addition, we design the fitness function using support, confidence and length of the rules for the purpose of extracting more practical and intelligible rules without predetermining the parameter thresholds. Besides, high-dimensional problems can also be solved using our algorithm. The experiments using substantial flight data are conducted to illustrate the excellent performance of the proposed BSO algorithm comparing to the Apriori algorithm and Genetic algorithm (GA). Furthermore, the classification problems with two datasets from UCI database are also used to verify the practicability and universality of the proposed method based on BSO algorithm.