Research And Application On Learning Algorithm Of Autonomic Learning Subspace Based On DFL

Dynamic fuzzy logic theory is a new and effective theory when handling dynamic, imprecise, uncertain or vague information. Autonomic learning has always been an important issue which is concerned by both educationalists and psychologists, and it is also a hot issue of machine learning research. Autonomic learning is not only dynamic but also fuzzy. Therefore, this paper chooses dynamic fuzzy logic to study autonomic learning, and has achieved the following several achievements after nearly three years of efforts:(1) Proposed the axiomatic system of autonomic learning subspace and given reasonable interpretation, in order to formally describe autonomic learning system.(2) Introduced the model of autonomic learning based on DFL.(3) The learning algorithm of autonomic learning subspace based on DFL was presented, which further enriched the contents of autonomic learning.(4) Used the learning algorithm of autonomic learning subspace based on DFL in Parameter tuning of PID, and has achieved satisfactory results.This paper studied the model and learning algorithm of autonomic learning based on dynamic fuzzy logic, thus provided the theoretical foundation to solve the dynamic fuzzy problem of autonomic learning system. However, all the work is preliminary and much of them need advanced research. As future work, we plan to improve the learning algorithm and to apply the autonomic learning model to more fields, and so on.