Neural network with fuzzy dynamic logic

The paper describes a neural network utilizing models and extending fuzzy logic. The motivation is to explain the process of learning as joint model improvement and fuzziness reduction. An initial state of this neural network is highly fuzzy with uncertain knowledge; it dynamically evolves into a low-fuzzy state of certain knowledge. This neural system resembles several known mechanisms of the mind and overcomes certain long-standing difficulties in several application fields. We present an example and briefly discuss mechanisms of concepts, emotions, including aesthetic emotions, instincts, conscious, unconscious, imagination, perception, cognition and relate them to the introduced neural network and the novel mathematics of dynamic logic.