Recursive Fuzzy Modeling Based on Fuzzy Interpolation

Abstract

This paper introduces a new fuzzy modeling of an unknown system. The heart of the proposed modeling is fuzzy interpolation involving resolution reduction that generates two different types of information to define single-input, single-output subsystems for an unknown system. Input is identified using a heuristic based on the proposed technique. System behavior is defined as a curve corresponding to individual input. These curves are found using fuzzy curve fitting applied to data points. Behavior curves are saved one of three ways. When the first is geometrical, the second and third use a neural network as a linear prediction coder. Convergence of the linear prediction coder specifies the membership function domain. Inference rules are extracted by observing the exactness of single-input behavior using a fuzzy method. The combination rule is new and combines single-input, single-output behavior to obtain the system model. Our proposal cancels noise well, recognizes actual input, and expresses complicated nonlinear systems with a very small number of rules.

Experimental Investigation of the Fault Tolerance of IDS Models

“…The ink drop spread (IDS) method is a modeling technique that is proposed as a new approach to soft computing [6] [7]. This method is analogous to fuzzy logic since it is algorithmically modeled on the information-handling processes of the human brain.…”

“…x (iia2l (X2)Yll + wl2a22(X2)Y12 + w21all(Xl)Y21 + w22al2(Xl)Y22) (7) where Wik = fik (xi) denotes the weight of the narrow path of the kth IDS unit for the ith input variable; fik (.) represents the conversion function (6); and aij is the membership function of Aij.…”

Using IDS fitted Q to develop a real-time adaptive controller for dynamic resource provisioning in Cloud's virtualized environment

“…Thus, IDS properly suits the Qlearning problem that is an MISO problem (there is only Q-value as the output of approximation procedure). The IDS model is trained via a recursive partitioning algorithm, called the active learning method (ALM) [35]. This paper proposes a novel fitted Q method to accelerate the modeling procedure inside Q-learning for continuous RL problems.…”

Active Learning Method for the Determination of Coupling Factor and External Q in Microstrip Filter Design

“…The time consuming process of ANN training and difficult interpretation of the knowledge embedded in the trained ANNs in a comprehensive form decreases the interest of using these modeling techniques if there is any alternative. ALM which is a relatively new soft computing technique does not suffer from the mathematical complexity of fuzzy algorithms and the difficult interpretability of ANN-based techniques [46,47]. It has an intelligent informationhandling process such as human brain and can be interpreted as a recursive fuzzy method which can express any multi-input singleoutput (MISO) system as the combination of some single-input singleoutput (SISO) one.…”