Adaptive Classification of Big Data Flight Sample

For the classification of single flight health monitoring of big data samples, we propose a correlation coefficient of membership classification methods. The method is based on the self-organization of fair competition algorithm optimization (KCN) combined with fuzzy k-means (FKM) neural network, the network will give a crude clustering KCN class center and class number as the root into the refinement FKM processing, thereby improving final accuracy of clustering results; with fairness algorithm treated separately ganglion win rate in order to improve the utilization of neurons. The classification method maintains the nature of the flight characteristics of the sample, the correlation coefficient through membership classification algorithms to solve the resampling flight sample classification problems. Can be updated in real time adaptive flight quickly and accurately classify samples for subsequent flight loads to improve the prediction accuracy of the simulation results demonstrate the feasibility, effectiveness of the method.