Modified Generalized Discriminant Analysis For Radar HRRP Recognition

Generalized discriminant analysis (GDA) is a nonlinear extension of the classical linear discriminant analysis (LDA) via kernel trick. As a feature extraction method, it has been proven successful in many applications such as radar high-range-resolution profiles (HRRP) recognition. However, GDA often suffers from the so-called small sample size problem (SSS) which exists in high-dimensional pattern recognition data. To overcome this weakness, we present a new algorithm for solving GDA by utilizing the idea of direct-LDA (DLDA) in this paper. Experiments based on three measured airplanes data are conducted to evaluate the effectiveness of the proposed method. From the results we can see that the new algorithm is more transparent and easier to be implemented than the traditional one, while keeping competitive classification accuracy.