This paper introduces a new structure in neural networks called TD-(CMAC), an extension to the conventional CEREBELLAR MODEL ARITHMETIC COMPUTER ((CMAC)), having reasonable ability in time series prediction. TD-(CMAC), the conventional (CMAC) and a classical neural network MODEL called Multi-Layer Perceptron (MLP) are simulated and evaluated for 1-hour-ahead prediction and 24-hour-ahead prediction of carbon monoxide as one of primary air pollutants. Carbon monoxide data used in this evaluation were recorded and averaged at Villa station in Tehran, Iran from October 3thrd. 2001 to March 14th. 2002 at one-hour intervals. The results show that the errors made by TD-(CMAC) is fewer than those made by other MODELs.

This paper introduces a new structure in neural networks called as TD-(CMAC) (Time Delay CEREBELLAR MODEL Articulation Controller), an extension to the conventional (CMAC).,TD-(CMAC) has reasonable ability in time series prediction and requires fewer memory sizes. The suggested new MODEL, the conventional (CMAC) and the classical multi-layer perceptron neural network (MLP) are demonstrated and validated for one-day-ahead prediction of daily minimum/maximum temperature. The comparison between different MODELs shows that the TD-(CMAC) is likely to be suitable for our purpose.      

CEREBELLAR MODEL Articulation Controller Neural Network is a computational MODEL of cerebellum which acts as a lookup table. The advantages of (CMAC) are fast learning convergence, and capability of mapping nonlinear functions due to its local generalization of weight updating, single structure and easy processing. In the training phase, the disadvantage of some (CMAC) MODELs is unstable phenomenon or slower convergence speed due to larger fixed or smaller fixed learning rate respectively. The present research deals with offering two solutions for this problem. The original idea of the present research is using changeable learning rate at each state of training phase in the (CMAC) MODEL. The first algorithm deals with a new learning rate based on reviation of learning rate. The second algorithm deals with number of training iteration and performance learning, with respect to this fact that error is compatible with inverse training time. Simulation results show that this algorithms have faster convergence and better performance in comparison to conventional (CMAC) MODEL in all training cycles.

A control strategy on a hybrid vehicle can be implemented through different methods. In this paper, the CEREBELLAR MODEL Articulation Controller ((CMAC)) and Radial Basis Function (RBF) neural networks were applied to develop an optimal control strategy for a split parallel hydraulic hybrid vehicle. These networks contain a nonlinear mapping, and, also, the fast learning procedure has made them desirable for online control. The RBF network was constructed with the use of the K-mean clustering method, and the (CMAC) network was investigated for different association factors. Results show that the binary (CMAC) has better performance over the RBF network. Also, the hybridization of the vehicle results in considerable reduction in fuel consumption.

Data Envelopment Analysis (DEA) is a nonparametric approach for evaluating the relative efficiency of a homogenous set of Decision Making Units (DMUs). To evaluate the relative efficiency of all DMUs, DEA MODEL should be solved once for each DMU. Therefore, by increasing the number of DMUs, computational requirements are increased. The CEREBELLAR MODEL Articulation Controller ((CMAC)) is a neural network that resembles a part of the brain known as cerebellum. The (CMAC) network with a simple structure is capable of estimating nonlinear functions, system MODELling and pattern recognition. Meanwhile, the (CMAC) approach has fast learning convergence and local generalization in comparison to other networks. The present paper is concerned with assessing the efficiency of DMUs by the (CMAC) neural network for the first time. The proposed approach is applied to a large set of 600 Iranian bank branches. The efficiency results are analyzed and compared with the Multi-layer Perceptrons (MLP) network outcomes. Based on the results, it can be seen that the DEA-(CMAC) results tend to be similar to those of DEA-MLP in terms of accuracy. In addition, the Mean Squared Error (MSE) in DEA-(CMAC) decreases much faster than that in DEA-MLP. The DEA-(CMAC) MODEL takes 1008 and 1107 iterations to reach MSE errors of 2.03×10-4  and of 6.01×10-4 , respectively, while the DEA-MLP MODEL takes 1190 iterations keeping the MSE error stable at 2.07×10-1 . Moreover, DEA-(CMAC) requirements for CPU time are far less than those needed by DEA-MLP.