FUZZY DYNAMIC LOGIC

Fuzzy logic is extended toward dynamic adaptation of the degree of fuzziness. The motivation is to explain the process of learning as a joint model improvement and fuzziness reduction. A learning system with fuzzy models is introduced. Initially, the system is in a highly fuzzy state of uncertain knowledge, and it dynamically evolves into a low-fuzzy state of certain knowledge. We present an image recognition example of patterns below clutter. The paper discusses relationships to formal logic, fuzzy logic, complexity and draws tentative connections to Aristotelian theory of forms and working of the mind.