In recent years, extensive research efforts have been made to search for substellar companions of nearby stars, and a number of approaches have been proposed to address this problem, including astronometric and velocity monitoring, coronagraphic imaging, and infrared search.
The proposed research activity is an extension of our previous research on 10-micron search for cool companions of nearby stars. A big challenge of the direct detection method is the difficulty of distinguishing between any faint object that may be present and artifacts or noise in the images. To address this challenge, we propose to employ support vector machine (SVM), a widely used tool in pattern classification field due to its robust and excellent performance, to improve the accuracy of cool companion detection. In this multidisciplinary approach, we investigate the performance of a technique used in electrical engineering to an astronomical research question.