Practicality of Modeling Systems Using the IDS Method: Performance Investigation and Hardware Implementation
真之 村上 · 2008
The performance characteristics robustness, speed, and tractability are important for the realization of practical computing systems. The concept of soft computing enables the achievement of such practical characteristics by tolerating imprecision and uncertainty instead of depending on exact mathematical computations. The ink drop spread (IDS) method is a modeling technique that has been proposed as a new approach to soft computing. This method is characterized by a modeling process that uses image information without including complex formulas. The model structure is a parallel distributed system comprising multiple SISO units, each of which independently functions as a modeling engine; each SISO unit is termed an “IDS unit.” Each IDS unit extracts a feature of the modeling target in the form of an image. The upper processing unit performs inference by aggregating the image information from all the IDS units. The objective of this dissertation is to investigate the performance of the IDS method as a soft computing tool and to reveal the practicality of this method. First, the modeling ability of the IDS method is presented by using three benchmarks concerning logic operation, function approximation, and classification. Next, the performance of the IDS method is investigated in terms of robustness, speed, and tractability, which are typical criteria that determine the importance of soft computing tools. The robustness is evaluated on the basis of noise tolerance and fault tolerance. Based on these criteria, we compare the IDS models with artificial neural networks (ANNs) and fuzzy inference systems (FISs). In several studies on ANNs, Hwang’s five-function set is used as the regression benchmark, and the generalization performance is evaluated for each set of noiseless and noisy training data. By using this benchmark, we present the noise tolerances of the IDS models, multilayer perceptrons (MLPs), radial basis function networks (RBFNs), and adaptive neuro-fuzzy inference systems (ANFISs). The MLP is the most popular model of neural networks. The RBFN is a variant of the ANN and is known to have superior fault tolerance and fast convergence. The ANFIS is characterized by its hybrid learning; the parameters of the premise part and consequent part in its fuzzy inference are adjusted by the gradient descent method and least squares method, respectively. The structure of IDS models is similar to that of ANNs: they comprise distributed processing units. The implementation of such parallel processing networks with dedicated hardware provides fault tolerance. We evaluate the fault tolerances of IDS models and ANNs for single unit failure using the stuck-at fault model. Most studies on the fault tolerance of ANNs investigate fault tolerance using classification tasks. This is because in the approximation of continuous functions, the failure of a single hidden unit causes considerable degradation in computation. Our experiments use function approximations that are more difficult than