Faint spatial object classifier construction based on data mining technology
Xin Lou, Yang Zhao, Yurong Liao, Yongming Nie · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2016
Data mining can effectively obtain the faint spatial object’s patterns and characteristics, the universal relations and other implicated data characteristics, the key of which is classifier construction. Faint spatial object classifier construction with spatial data mining technology for faint spatial target detection is proposed based on theoretical analysis of design procedures and guidelines in detail. For the one-sidedness weakness during dealing with the fuzziness and randomness using this method, cloud modal classifier is proposed. Simulating analyzing results indicate that this method can realize classification quickly through feature combination and effectively resolve the one-sidedness weakness problem.