Assessing Inductive Logic Programming Classification Quality by Image Synthesis

One of the major difficulties in classification is the assessment of classification quality. This paper builds on a recently developed classification technique, based on Inductive Logic Programming to propose a way of visually reconstructing the learnt classes in order to give a feedback to the user. The system translates positioning constraints into a linear programming problem, which can be solved with standard state-of-theart approaches. The result can then be used for generating a synthetic image, representative of the class.