Explainable clustering has become increasingly important as it presents the clustering results in a manner that can be easily understood by the end-users. However, most existing explainable clustering approaches generate only a single explanation such as a decision tree for the given clustering result, overlooking that multiple valid interpretations may exist. To address this limitation, we propose the CME (Clustering with Multiview Explanations) algorithm. This method constructs a diverse set of candidate decision trees by retaining multiple splitting points at each node, where each splitting point is valid in a statistical sense. The tree similarity based on optimal path matches across corresponding clusters is then used to build a similarity graph. A subset of representative decision trees are selected by solving the minimum dominating set problem defined on the corresponding similarity graph. Experiments on 10 real-world categorical datasets demonstrate that CME provides multiple complementary explanations that may be missed by existing algorithms. Moreover, these additional decision trees can be more accurate and interpretable than those ones identified by baseline methods.

Ultrawideband synthetic aperture radar (SAR) (UWB SAR) is a sufficient approach to detecting land mines over large areas from a safe standoff distance. UWB SAR detection involves three steps: imaging, prescreening, and discrimination. Many studies have concentrated on imaging and discrimination, whereas few have highlighted prescreening. We propose a novel prescreening method for extracting the region of interest from an entire SAR image. The proposed method is based on feature point matching in contrast to traditional approaches, which are based on the constant false alarm rate (CFAR) technique. Our method incorporates knowledge on the electromagnetic scattering caused by land mines into the prescreening process and does not rely on clutter environments. Thus, the method more efficiently eliminates nonhomogeneous clutter than do CFAR-based methods and is more advantageous in processing large-area SAR images. The efficiency of the proposed method is validated by real data on an airship-mounted UWB SAR system.

A vehicle-mounted forward-looking ground-penetrating radar (GPR) with multiple-input and multiple-output (MIMO) array can obtain the high-resolution image of its front area to perform the standoff detection of landmines. The major challenge for the GPR landmine detection over wide areas is the very high false alarm rate when maintaining a high detection probability. In this paper, a novel feature extraction method is proposed to obtain the bistatic scattering information from the MIMO array image to discriminate landmines from clutter. To realize the goal, an imaging model of the MIMO array is firstly developed. Based on the imaging model, the bistatic scattering function of a suspected object is estimated from its MIMO array image using the space-wavenumber processing. Images of different incident angles and bistatic angles at some resonance frequencies are selected from the estimated bistatic scattering function to represent the scattering characteristics. In order to obtain the scale, rotation, and translation invariant feature vector, Hu moment invariants of the selected images are calculated to form the low-dimensional feature vector. The experimental results show that the proposed method can offer an efficient feature vector for the landmine discriminator to improve the detection performance.