Autonomous mental development for algorithm recognition

Algorithm recognition is concerned with program understanding. In the past decades, several approaches have been studied in this area, but most of them are based on a library where predefined templates are stored. Such template-based approaches encounter an obstacle that it is difficult to know how many templates are required to recognize a given algorithm in advance. To avoid this obstacle, we apply the idea of autonomous mental development (AMD) to algorithm recognition. In our approach, vectors with initially randomized values will be developed autonomously into vectorial templates suitable for algorithm recognition. Our experiment illustrates how the vectorial templates are grown up. The result shows that our method could achieve as high as 93.4% recognition accuracy in average.