作者: Justin Metcalf , Shannon D. Blunt , Braham Himed
DOI: 10.1109/RADAR.2015.7131215
关键词:
摘要: We consider the requirements of cognitive radar detection in presence non-Gaussian clutter. A pair machine learning approaches based on non-linear transformations order statistics are examined with goal adaptively determining optimal threshold within low sample support regime. The impact these algorithms false alarm rate is also considered. It demonstrated that adaptive estimate effective even when distribution question unknown to algorithm.