Is readability compatible with accuracy?: from Neuro-Fuzzy to Lazy Learning

The composition of simple local models for approximating complex nonlinear mappings is a common practice in recent modeling and control literature. This paper presents a comparative analysis of two different local approaches: the neuro-fuzzy inference system and the lazy learning approach. A neuro-fuzzy system is an hybrid representation which combines the linguistic description of fuzzy inference systems with learning procedures inspired by neural networks. Lazy learning is a memory-based technique that uses a query-based approach to select the best local model configuration by assessing and comparing different alternatives in cross-validation. The two approches are compared both as learning algorithms and as identification modules of an adaptive control system. The paper will show how the lazy learning is able to provide better modeling accuracy and control performance at the cost of a reduced readibility of the resulting approximator. Illustrative examples of identification and control of a nonlinear system starting from simulated data are given.