Cerebellar Model Articulation Controller

This paper introduces an effective intelligent controller for robotic systems with uncertainties. The proposed method is a novel self-organizing fuzzy cerebellar model articulation controller (NSOFC) which is a combination of a cerebellar model articulation controller (CMAC) and sliding mode control (SMC). We also present a new Gaussian membership function (GMF) that is designed by the combination of the prior and current GMF for each layer of CMAC. In addition, the relevant data of the prior GMF is used to check tracking errors more accurately. The inputs of the proposed controller can be mixed simultaneously between the prior and current states according to the corresponding errors. Moreover, the controller uses a self-organizing approach which can increase or decrease the number of layers, therefore the structures of NSOFC can be adjusted automatically. The proposed method consists of a NSOFC controller and a compensation controller. The NSOFC controller is used to estimate the i...

Cerebellar model articulation controller (CMAC) neural network is a real alternative to MLP and RBF and has some advantageous features: its training is fast and its architecture is especially suitable for digital hardware implementation. The price of these attractive features is its rather poor Capability. The poor capability can be improved if a recently proposed regularization andlor a kernel version of the CMAC is used. The paper presents the first results of a study about the efticient hardware realization of various versions of CMAC. Three versions have been developed the classical Albus CMAC, its regularized version, which has better generalization capability and a kernel CMAC. The solutions, that were developed using high-level synthesis (Handel C), are based on Virtex FPGA.

A data stream is an ordered sequence of instances that can be read only once or a small number of times using limited computing and storage capabilities. Stream data analysis is a critical issue in many application areas such as network fraud detection, stock market prediction, and web searches. In this research, our previously proposed FCMAC-BYY, that uses Bayesian Ying-Yang (BYY) learning in the fuzzy cerebellar model articulation controller (FCMAC), will be advanced by evolutionary computation and dynamic rule construction. The developed FCMAC-EBYY has been applied to a real-time stream data analysis problem of traffic flow prediction. The experimental results illustrate that FCMAC-EBYY is indeed capable of producing better performance than other representative neuro-fuzzy systems.

T his paper describes the design, im plem entation and perform ance evaluations o f the application developed for real-tim e drinking w ater quality detection using the C erebellar M odel A rticulation C o n tro ller (CM A C) Artificial N eural N etw orks (A N N s). M o st drinking water systems currently m o n ito r a significant n u m b er o f w ater quality param eters at the plant. T hese are required for com pliance and m aintenance o f water quality as the w ater enters the distribution system. T h e threat o f chemical o r microbiological contam ination o f drinking w ater is obvious, and w ould be an effective way o f causing devastating public health consequences. In this study we propose a practical approach for the detection and classification o f chemical and microbiological contam inants in drinking water. C M A C A N N is intended for the use in the detection process o f drinking w ater quality w here real-tim e capabilities o f the netw ork are o f prim e im portance. C M A C has been chosen as an alternative to M ultilayer Perceptron (M LP) because o f its exceptional learning capability. T h e learning speed o f the C M A C A N N algorithm is m u ch higher than that o f M LP using a backpropagation (BP) algorithm. T h e detection (rate) o f the n ew w ater samples has also been so fast (high) and accurate that the rate o f success alm ost turn ed o u t to be 100% as com pared to an ordinary M LP using the BP algorithm o f w hich the success rate was found to be 98%.

Electronic noses are being developed as systems for the automated detection and classification of odors, vapors, and gases. Artificial neural networks (ANNs) have been used to analyze complex data and to recognize patterns, and have shown promising results in recognition of volatile compounds and odors in electronic nose applications. When an ANN is combined with a sensor array, the number of detectable chemicals is generally greater than the number of unique sensor types. The odor sensing system should be extended to new areas since its standard style where the output pattern from multiple sensors with partially overlapped specificity is recognized by a neural network or multivariate analysis. This paper describes the design, implementation and performance evaluations of the application developed for hazardous odor recognition using Cerebellar Model Articulation Controller (CMAC) based neural networks.

Konveksna optimizacija u u~enju CMAC neuronskih mre`a.. Jednostavnost gra|e i algoritama u~enja od iznimne su va`nosti u primjenama neuronskih mre`a u stvarnom vremenu. CMAC neuronska mre`a s asocijativnom memorijskom organizacijom i Hebbianovim algoritmom u~enja udovoljava ovim zahtjevima. Me|utim, Hebbianov algoritam u~enja ne daje dobre rezultate pri off-line identifikaciji, koja se koristi kao pripremna faza za on-line identifikaciju. U ovom se~lanku pokazuje da se optimalne vrijednosti parametara CMAC neuronske mre`e mogu dobiti primjenom tehnika konveksne optimizacije. Za standardnu l 2 aproksimaciju koristi se kvadratno programiranje (QP), a za l 1 i l ∞ aproksimacije linearno programiranje (LP). U oba je slu~aja jednostavno uklju~iti fizikalna ograni~enja na vrijednosti parametara u algoritam optimizacije.

In this paper, a systematic treatment for developing a noise removal system based on the fundamental principle of reinforcement learning and fuzzy cerebellar model articulation controller (FCMAC) is presented. The proposed system improves its performance over time through two mechanisms. First, the modified stochastic real-valued algorithm, learning from its own mistakes via the reinforcement signal and reinforcing its action to improve future performance, is used for searching the optimal noise spectrum for the overall training system. Second, system states associated with the positive reinforcement are memorized by FCMAC-based neurons, where, in the future, similar states will share the experiences already stored there and then lead the action to a more positive situation. In this work, FCMAC’s intrinsically poor approximation of rapidly varying functions is solved by taking the complex semicepstrum. In addition, the FCMAC provides an improvement in accuracy of function approximat...

A neuro-fuzzy system which is embedded in the conventional control theory is proposed to tackle physical learning control problems in this paper. The control scheme is composed of two elements. The first element, the fuzzy sliding mode controller (FSMC), is used to drive the state variables to a specific switching hyperplane or a desired trajectory. The second one is developed based on the concept of the self-organizing fuzzy cerebellar model articulation controller (FCMAC) and adaptive heuristic critic (AHC). Both compose a forward compensator to reduce the chattering effect or cancel the influence of system uncertainties. A geometrical explanation on how the FCMAC algorithm works is provided and some refined procedures of the AHC are presented as well. Simulations on smooth motion of three-link robot is given to illustrate the performance and applicability of the proposed control scheme.

We present a Cerebellar Model Articulation Controller (CMAC) neural network architecture that is suitable for continuous learning, by providing stability and plasticity to the network, which is difficult to achieve in the majority of the neural networks. We also present two learning algorithms, namely, (a) a learning process and (b) a re-learning process. The learning process involves transforming an input into a set of codes such that similar inputs are coded as overlapping codes, while dissimilar inputs are coded as disjoint codes. Each category is represented by a set of codes stored in a memory as a part of the training. An input is classified to a category if the majority of its codes are in common with the codes corresponding to that category. In the re-learning process, we extract the statistical information about each category in terms of the frequency of appearance of a code and the number of signals classified to the category. Based on this statistical information, new exemplars (ideal signals representing each category) for each category are created. Based on these new exemplars the network is trained using the learning process. The trained net is then used to classify ultrasonic signals obtained from a damaged sample. The classification results are compared with the classification results obtained from the standard C-scan imaging techniques and it is found that the classification by the CMAC system with a re-learning algorithm is 92% accurate.

This paper proposes a recurrent fuzzy cerebellar model articulation controller (RFCMAC) for a dual-axis micromotion stage powered by piezoelectric actuators. The proposed RFCMAC employs an adjustable input space quantization method with a learning rate capable of self-tuning. The proposed RFCMAC is aimed at dealing with the adverse effects due to unknown hysteresis nonlinearity, modeling uncertainty, and external disturbance that are commonly found in the piezoelectric actuator systems. The proposed RFCMAC mainly focuses on overcoming the drawbacks of the conventional CMAC control scheme when used to control a piezoelectric actuator system. In particular, the adjustable input space quantization method uses a simple repartitioning decision function to determine an appropriate input space quantization. In addition, the self-tuning law for the learning rate of the proposed RFCMAC is derived based on the discretetime Lyapunov function so that convergence and faster learning can be guaranteed. Finally, in order to validate the effectiveness of the proposed control methodology, several contour-following experiments were conducted on a dual-axis piezoelectric actuated micromotion stage. Experimental results show that the proposed control scheme is feasible, and its performance is superior to that of the conventional CMAC control scheme.

In this paper, a novel cerebellar model articulation controller (CMAC)-Based compensator is proposed to limit bound required in supervisory control systems. There are two structures in the proposed schemes: one is supervisory controller and the other is the CMAC-Based compensator. The supervisory controller can ensure Lyapunov stability of the controlled system in the presence of significant plant uncertainties, if the perfect control is estimated. The CMAC is employed to learn the perfect control, but a model error will exist in the learning process. The object of CMAC-based compensator is to suppress this model error, so that the supervisory of can be rationalized for uncertain nonlinear systems. Finally, simulation results demonstrate that the CMAC-based compensator not only can limit the bound required in supervisory controllers, but also can significantly improve the control performance.

In this study, a novel indirect adaptive controller is introduced for a class of unknown nonlinear systems. The proposed method provides a simple control architecture that merges from the cerebellar model articulation controller (CMAC) network and hierarchical fuzzy logic; therefore, the complicated CMAC structure can be simplified. The overall adaptive scheme guarantees the uniform stability of the closed-loop system. A simulation is presented to demonstrate the performance of the proposed methodology.

A digital Video Stabilization (DVS) system removes the unwanted shaking in the videos acquired by hand-held cameras and preserves the panning. In this paper, a digital video stabilization system is proposed based upon adaptive cerebellar model articulation controller (CMAC) filtering. A CMAC is a manifestation of the associative memory learning structure present in the cerebellum of human being. Adaptive CMAC filtering has favorable properties of small size, good generalization, rapid learning and dynamic response. Thus, it is more suitable for high-speed signal processing applications. The adaptive CMAC is used to adjust the coefficients of IIR filter employed in the proposed model. The training of CMAC is based upon fuzzy rule. The efficacy of the proposed adaptive CMAC filtering has been validated by evaluating it on a set of test video sequences.

The timely and accurate detection of computer and network system intrusions has always been an elusive goal for system administrators and information security researchers. Existing intrusion detection approaches require either manual coding of new attacks in expert systems or the complete retraining of a neural network to improve analysis or learn new attacks. This paper presents a new approach to applying adaptive neural networks to intrusion detection that is capable of autonomously learning new attacks rapidly using feedback from the protected system.

This paper proposes a self-organizing control system for uncertain nonlinear systems. The proposed neural network is composed of a conventional brain emotional learning network (BEL) and a cerebellar model articulation controller network (CMAC). The input value of the network is feed to a BEL channel and a CMAC channel. The output of the network is generated by the comprehensive action of the two channels. The structure of the network is dynamic, using a self-organizing algorithm allows increasing or decreasing weight layers. The parameters of the proposed network are on-line tuned by the brain emotional learning rules; the updating rules of CMAC and the robust controller are derived from the Lyapunov function; in addition, stability analysis theory is used to guaranty the proposed controller's convergence. A simulated mobile robot is applied to prove the effectiveness of the proposed control system. By comparing with the performance of other neural-networkbased control systems, the proposed network produces better performance.

Konveksna optimizacija u u~enju CMAC neuronskih mre`a.. Jednostavnost gra|e i algoritama u~enja od iznimne su va`nosti u primjenama neuronskih mre`a u stvarnom vremenu. CMAC neuronska mre`a s asocijativnom memorijskom organizacijom i Hebbianovim algoritmom u~enja udovoljava ovim zahtjevima. Me|utim, Hebbianov algoritam u~enja ne daje dobre rezultate pri off-line identifikaciji, koja se koristi kao pripremna faza za on-line identifikaciju. U ovom se~lanku pokazuje da se optimalne vrijednosti parametara CMAC neuronske mre`e mogu dobiti primjenom tehnika konveksne optimizacije. Za standardnu l 2 aproksimaciju koristi se kvadratno programiranje (QP), a za l 1 i l ∞ aproksimacije linearno programiranje (LP). U oba je slu~aja jednostavno uklju~iti fizikalna ograni~enja na vrijednosti parametara u algoritam optimizacije.

Purpose – The purpose of this paper is to improve the flexibility and tracking errors of the controllers-based neural networks (NNs) for mobile manipulator robot (MMR) in the presence of time-varying uncertainties. Design/methodology/approach – The conventional backstepping force/motion control is developed by the wavelet fuzzy CMAC neural networks (WFCNNs) (for mobile-manipulator robot). The proposed WFCNNs are applied in the tracking-position-backstepping controller to deal with the uncertain dynamics of the controlled system. In addition, an adaptive robust compensator is proposed to eliminate the inevitable approximation errors, uncertain disturbances, and relax the requirement for prior knowledge of the controlled system. Besides, the position tracking controller, an adaptive robust constraint-force is also considered. The online-learning algorithms of the control parameters (WFCNNs, robust term and constraint-force controller) are obtained by using the Lyapunov stability theor...

High-precision trajectory control is considered as an important factor in the performance of industrial two-axis contour motion systems. This research presents an adaptive direct fuzzy cerebellar model articulation controller (CMAC) sliding mode control (DFCMACSMC) for the precise control of the industrial XY-axis motion system. The FCMAC was utilized to approximate an ideal controller, and the weights of FCMAC were on-line tuned by the derived adaptive law based on the Lyapunov criterion. With this derivation in mind, the asymptotic stability of the developed motion system could be guaranteed. The two-axis stage system was experimentally investigated using four contours, namely, circle, bowknot, heart, and star reference contours. The experimental results indicate that the proposed DFCMACSMC method achieved the improved tracking capability, and so reveal that the DFCMACSMC scheme outperformed other schemes of the model uncertainties and cross-coupling interference.

A digital Video Stabilization (DVS) system removes the unwanted shaking in the videos acquired by hand-held cameras and preserves the panning. In this paper, a digital video stabilization system is proposed based upon adaptive cerebellar model articulation controller (CMAC) filtering. A CMAC is a manifestation of the associative memory learning structure present in the cerebellum of human being. Adaptive CMAC filtering has favorable properties of small size, good generalization, rapid learning and dynamic response. Thus, it is more suitable for high-speed signal processing applications. The adaptive CMAC is used to adjust the coefficients of IIR filter employed in the proposed model. The training of CMAC is based upon fuzzy rule. The efficacy of the proposed adaptive CMAC filtering has been validated by evaluating it on a set of test video sequences.

T his paper describes the design, im plem entation and perform ance evaluations o f the application developed for real-tim e drinking w ater quality detection using the C erebellar M odel A rticulation C o n tro ller (CM A C) Artificial N eural N etw orks (A N N s). M o st drinking water systems currently m o n ito r a significant n u m b er o f w ater quality param eters at the plant. T hese are required for com pliance and m aintenance o f water quality as the w ater enters the distribution system. T h e threat o f chemical o r microbiological contam ination o f drinking w ater is obvious, and w ould be an effective way o f causing devastating public health consequences. In this study we propose a practical approach for the detection and classification o f chemical and microbiological contam inants in drinking water. C M A C A N N is intended for the use in the detection process o f drinking w ater quality w here real-tim e capabilities o f the netw ork are o f prim e im portance. C M A C has been chosen as an alternative to M ultilayer Perceptron (M LP) because o f its exceptional learning capability. T h e learning speed o f the C M A C A N N algorithm is m u ch higher than that o f M LP using a backpropagation (BP) algorithm. T h e detection (rate) o f the n ew w ater samples has also been so fast (high) and accurate that the rate o f success alm ost turn ed o u t to be 100% as com pared to an ordinary M LP using the BP algorithm o f w hich the success rate was found to be 98%.

Electronic noses are being developed as systems for the automated detection and classification of odors, vapors, and gases. Artificial neural networks (ANNs) have been used to analyze complex data and to recognize patterns, and have shown promising results in recognition of volatile compounds and odors in electronic nose applications. When an ANN is combined with a sensor array, the number of detectable chemicals is generally greater than the number of unique sensor types. The odor sensing system should be extended to new areas since its standard style where the output pattern from multiple sensors with partially overlapped specificity is recognized by a neural network or multivariate analysis. This paper describes the design, implementation and performance evaluations of the application developed for hazardous odor recognition using Cerebellar Model Articulation Controller (CMAC) based neural networks.

A key element in the solution of reinforcement learning problems is the value function. The purpose of this function is to measure the long-term utility or value of any given state. The function is important because an agent can use this measure to decide what to do next. A common problem in reinforcement learning when applied to systems having continuous states and action spaces is that the value function must operate with a domain consisting of real-valued variables, which means that it should be able to represent the value of infinitely many state and action pairs. For this reason, function approximators are used to represent the value function when a close-form solution of the optimal policy is not available. In this article, we extend a previously proposed reinforcement learning algorithm so that it can be used with function approximators that generalize the value of individual experiences across both state and action spaces. In particular, we discuss the benefits of using spar...

This paper presents an online learning control algorithm based on CMAC network to realize the tracking control problem of nonlinear system. Compared with the traditional two-dimensional input extended to multi-input multi-output nonlinear system, a new address space mapping method is introduced in the cerebellum controller to avoid collision of address space and improve the stability of the microcomputer image controller.

A macro structure cerebellar model articulation controller (CMAC) or MS_CMAC was developed by connecting several one-dimensional (1-D) CMACs as a tree structure, which decomposes a multidimensional problem into a set of 1-D subproblems, to reduce the computational complexity in multidimensional CMAC. Additionally, a trapezium scheme is proposed to assist MS_CMAC to model nonlinear systems. However, this trapezium scheme cannot perform a real smooth interpolation, and its working parameters are obtained through cross-validation. A quadratic splines scheme is developed herein to replace the trapezium scheme in MS_CMAC, named high-order MS_CMAC (HMS_CMAC). The quadratic splines scheme systematically transforms the stepwise weight contents of CMACs in MS_CMAC into smooth weight contents to perform the smooth outputs. Test results affirm that the HMS_CMAC has acceptable generalization in continuous function-mapping problems for nonoverlapping association in training instances. Nonoverlapping association in training instances not only significantly reduces the number of training instances needed, but also requires only one learning cycle in the learning stage. Index Terms-Cerebellar model articulation controller (CMAC), high-order MS_CMAC, macro structure cerebellar model articulation controller (MS_CMAC), quadratic splines.

This paper proposes a self-organizing control system for uncertain nonlinear systems. The proposed neural network is composed of a conventional brain emotional learning network (BEL) and a cerebellar model articulation controller network (CMAC). The input value of the network is feed to a BEL channel and a CMAC channel. The output of the network is generated by the comprehensive action of the two channels. The structure of the network is dynamic, using a self-organizing algorithm allows increasing or decreasing weight layers. The parameters of the proposed network are on-line tuned by the brain emotional learning rules; the updating rules of CMAC and the robust controller are derived from the Lyapunov function; in addition, stability analysis theory is used to guaranty the proposed controller's convergence. A simulated mobile robot is applied to prove the effectiveness of the proposed control system. By comparing with the performance of other neural-networkbased control systems, the proposed network produces better performance.

This paper proposes a self-organizing control system for uncertain nonlinear systems. The proposed neural network is composed of a conventional brain emotional learning network (BEL) and a cerebellar model articulation controller network (CMAC). The input value of the network is feed to a BEL channel and a CMAC channel. The output of the network is generated by the comprehensive action of the two channels. The structure of the network is dynamic, using a self-organizing algorithm allows increasing or decreasing weight layers. The parameters of the proposed network are on-line tuned by the brain emotional learning rules; the updating rules of CMAC and the robust controller are derived from the Lyapunov function; in addition, stability analysis theory is used to guaranty the proposed controller's convergence. A simulated mobile robot is applied to prove the effectiveness of the proposed control system. By comparing with the performance of other neural-networkbased control systems, the proposed network produces better performance.

Forecasting cash demands at automatic teller machines (ATMs) is challenging, due to the heteroskedastic nature of such time series. Conventional global learning computational intelligence (CI) models, with their generalized learning behaviors, may not capture the complex dynamics and time-varying characteristics of such real-life time series data efficiently. In this paper, we propose to use a novel local learning model of the pseudo self-evolving cerebellar model articulation controller (PSECMAC) associative memory network to produce accurate forecasts of ATM cash demands. As a computational model of the human cerebellum, our model can incorporate local learning to effectively model the complex dynamics of heteroskedastic time series. We evaluated the forecasting performance of our PSECMAC model against the performances of current established CI and regression models using the NN5 competition dataset of 111 empirical daily ATM cash withdrawal series. The evaluation results show that the forecasting capability of our PSECMAC model exceeds that of the benchmark local and global-learning based models. c

This paper presents a system based on a higher order Cerebeilar Model Articulation Controller to improve the linearity and to compensate the hysteresis of a control valve. To highlight the benefits and drawbacks of the proposed approach, a Masoneilan Camflex 11 valve is used as a practical test bead. These results show that CMAC can effectively improve the valve performance. This improvement can be obtained by first collecting a set of data and then allowing CMAC to learn the necessary actions to compensate the non-linear effects.

The rapidly increasing size of databases creates a need for new algorithms to solve multi-class categorisation problems. Machine learning techniques such as neural networks have been successfully applied to this class of problems. However training times for these techniques can blow out as the size of the database increases. Some of the desirable features of algorithms for large databases are low order time complexity, training with only a single pass of the data, and accountability for class assignment decisions. We propose a new training algorithm for Cerebellar Model Articulation Controller (CMAC) based classifiers, which possesses these features. The training algorithm proposed here is based on a kernel addition method. An empirical investigation of this training method has found it to be superior to traditional techniques both in accuracy and time required to learn mappings between input vectors and class labels.

Forecasting cash demands at automatic teller machines (ATMs) is challenging, due to the heteroskedastic nature of such time series. Conventional global learning computational intelligence (CI) models, with their generalized learning behaviors, may not capture the complex dynamics and time-varying characteristics of such real-life time series data efficiently. In this paper, we propose to use a novel local learning model of the pseudo self-evolving cerebellar model articulation controller (PSECMAC) associative memory network to produce accurate forecasts of ATM cash demands. As a computational model of the human cerebellum, our model can incorporate local learning to effectively model the complex dynamics of heteroskedastic time series. We evaluated the forecasting performance of our PSECMAC model against the performances of current established CI and regression models using the NN5 competition dataset of 111 empirical daily ATM cash withdrawal series. The evaluation results show that the forecasting capability of our PSECMAC model exceeds that of the benchmark local and global-learning based models. c

Forecasting cash demands at automatic teller machines (ATMs) is challenging, due to the heteroskedastic nature of such time series. Conventional global learning computational intelligence (CI) models, with their generalized learning behaviors, may not capture the complex dynamics and time-varying characteristics of such real-life time series data efficiently. In this paper, we propose to use a novel local learning model of the pseudo self-evolving cerebellar model articulation controller (PSECMAC) associative memory network to produce accurate forecasts of ATM cash demands. As a computational model of the human cerebellum, our model can incorporate local learning to effectively model the complex dynamics of heteroskedastic time series. We evaluated the forecasting performance of our PSECMAC model against the performances of current established CI and regression models using the NN5 competition dataset of 111 empirical daily ATM cash withdrawal series. The evaluation results show that the forecasting capability of our PSECMAC model exceeds that of the benchmark local and global-learning based models. c

This work presents the development of a Proportional Integral Derivative (PID) digital controller with gain planning based on (CMAC) Cerebellar Model Articulation Controller. This digital controller operates in closed loop and has a dynamic gain adjustment. The control strategy uses an algorithm that calculates the proportional, integral and derivative parcels. This algorithm provides the soft start requirements. Both, the gain planning and the soft start requirements, uses auxiliary variables to determine an appropriate and dynamic setup of the digital controller. These auxiliary variables impose several restrictions on the digital controller and this one must have adaptive characteristics. The artificial neural network is used to estimate these auxiliary variables. The learning process depends on the data acquisition or the mathematical model simulations. The digital controller can operate in real-time conditions and the sample frequency can reach 10kHz.

A macro structure cerebellar model articulation controller (CMAC) or MS_CMAC was developed by connecting several one-dimensional (1-D) CMACs as a tree structure, which decomposes a multidimensional problem into a set of 1-D subproblems, to reduce the computational complexity in multidimensional CMAC. Additionally, a trapezium scheme is proposed to assist MS_CMAC to model nonlinear systems. However, this trapezium scheme cannot perform a real smooth interpolation, and its working parameters are obtained through cross-validation. A quadratic splines scheme is developed herein to replace the trapezium scheme in MS_CMAC, named high-order MS_CMAC (HMS_CMAC). The quadratic splines scheme systematically transforms the stepwise weight contents of CMACs in MS_CMAC into smooth weight contents to perform the smooth outputs. Test results affirm that the HMS_CMAC has acceptable generalization in continuous function-mapping problems for nonoverlapping association in training instances. Nonoverlapping association in training instances not only significantly reduces the number of training instances needed, but also requires only one learning cycle in the learning stage. Index Terms-Cerebellar model articulation controller (CMAC), high-order MS_CMAC, macro structure cerebellar model articulation controller (MS_CMAC), quadratic splines.

The rapidly increasing size of databases creates a need for new algorithms to solve multi-class categorisation problems. Machine learning techniques such as neural networks have been successfully applied to this class of problems. However training times for these techniques can blow out as the size of the database increases. Some of the desirable features of algorithms for large databases are low order time complexity, training with only a single pass of the data, and accountability for class assignment decisions. We propose a new training algorithm for Cerebellar Model Articulation Controller (CMAC) based classifiers, which possesses these features. The training algorithm proposed here is based on a kernel addition method. An empirical investigation of this training method has found it to be superior to traditional techniques both in accuracy and time required to learn mappings between input vectors and class labels.

Liver performs several numbers of metabolic functions that are essential to human life. These functions make the liver one of the most important organs in the human body. There are diseases that occur in the liver in short time (acute) and long time (chronic). These diseases could occur because of medications, alcohol, viruses, or excessive fat accumulation or deposit in the liver. Some of these diseases are the inflammation of the liver, insufficient liver performance, Hepatitis A, B, C, D, and liver cirrhosis. If the liver malfunctions in anyway, people know that they are putting their life at risk. For this reason, diagnosing any disease in the liver is important and sometimes difficult. It is also important to notice the diagnosis of the patient at an early stage as the symptoms arise so that the patient might be able to carry on a normal life. The objective of this article is to diagnose the liver disease using an application of the CMAC (Cerebellar Model Articulation Controlle...

Konveksna optimizacija u u~enju CMAC neuronskih mre`a.. Jednostavnost gra|e i algoritama u~enja od iznimne su va`nosti u primjenama neuronskih mre`a u stvarnom vremenu. CMAC neuronska mre`a s asocijativnom memorijskom organizacijom i Hebbianovim algoritmom u~enja udovoljava ovim zahtjevima. Me|utim, Hebbianov algoritam u~enja ne daje dobre rezultate pri off-line identifikaciji, koja se koristi kao pripremna faza za on-line identifikaciju. U ovom se~lanku pokazuje da se optimalne vrijednosti parametara CMAC neuronske mre`e mogu dobiti primjenom tehnika konveksne optimizacije. Za standardnu l 2 aproksimaciju koristi se kvadratno programiranje (QP), a za l 1 i l ∞ aproksimacije linearno programiranje (LP). U oba je slu~aja jednostavno uklju~iti fizikalna ograni~enja na vrijednosti parametara u algoritam optimizacije.

Traffic prediction is a critical element in traffic control today. With the increase of transportation, an effective traffic prediction allows to prevent traffic problems. This research aims to propose a novel approach to traffic prediction using Ying-Yang Fuzzy Cerebellar Model Articulation Controller (YY-FCMAC). The model is motivated from the famous Chinese ancient Ying-Yang philosophy, which views everything as a product of conflict-harmony process between Ying and Yang. That principle is applied to find the optimal number of clusters and fuzzy sets in the fuzzification phase of the hybrid fuzzy-neural YY-FCMAC network. The analyzed experiment on a set of real traffic data flow of the eastbound Pan Island Expressway (PIE) in Singapore shows the effectiveness of the YY-FCMAC in universal approximation and prediction.

Cerebellar Model Articulation Controller (CMAC) NN is a computational model of cerebellum introduced as an alternative to backpropagated multilayer networks to control robot arms. From then it has seen many improvements and has been applied in many other areas as a general NN. These improvements have been in the context of generalization, learning techniques, differentiability, memory size, fuzzification and hardware implementation. This paper is a systematic review of CMAC's different structures and applications.

Cerebellar Model Articulation Controller (CMAC) NN is a computational model of cerebellum introduced as an alternative to backpropagated multilayer networks to control robot arms. From then it has seen many improvements and has been applied in many other areas as a general NN. These improvements have been in the context of generalization, learning techniques, differentiability, memory size, fuzzification and hardware implementation. This paper is a systematic review of CMAC's different structures and applications.

Cerebellar Model Articulation Controller (CMAC) NN is a computational model of cerebellum introduced as an alternative to backpropagated multilayer networks to control robot arms. From then it has seen many improvements and has been applied in many other areas as a general NN. These improvements have been in the context of generalization, learning techniques, differentiability, memory size, fuzzification and hardware implementation. This paper is a systematic review of CMAC's different structures and applications.

The cerebellum is a brain region important for a number of motor and cognitive functions. It is able to generate error correction signals to drive learning and for the acquisition of memory skills. The cerebellar model articulation controller (CMAC) is a neural network inspired by the neurophysiologic theory of the cerebellum and is recognized for its localized generalization and rapid algorithmic computation capabilities. The main deficiencies in the basic CMAC structure are: 1) it is difficult to interpret the internal operations of the CMAC network and 2) the resolution (quantization) problem arising from the partitioning of the input training space. These limitations lead to the synthesis of a fuzzy quantization technique and the mapping of a fuzzy inference scheme onto the CMAC structure. The discrete incremental clustering (DIC) technique is employed to alleviate the quantization problem in the CMAC structure, resulting in the fuzzy CMAC (FCMAC) network. The Yager inference scheme (Yager), which possesses firm fuzzy logic foundation and maps closely to the logical implication operations in the classical (binary) logic framework, is subsequently mapped onto the FCMAC structure. This results in a novel fuzzy neural architecture known as the fuzzy cerebellar model articulation controller-Yager (FCMAC-Yager) system. The proposed FCMAC-Yager network exhibits learning and memory capabilities of the cerebellum through the CMAC structure while emulating the human way of reasoning through the Yager. The new FCMAC-Yager network employs a two-phase training algorithm consisting of structural learning based on the DIC technique and parameter learning using hebbian learning (associative long-term potentiation). The proposed FCMAC-Yager architecture is evaluated using an extensive suite of real-life applications such as highway traffic-trend modeling and prediction and performing as an early warning system for bank failure classification and medical diagnosis of breast cancer. The experimental results are encouraging.

Traffic prediction is a critical element in traffic control today. With the increase of transportation, an effective traffic prediction allows to prevent traffic problems. This research aims to propose a novel approach to traffic prediction using Ying-Yang Fuzzy Cerebellar Model Articulation Controller (YY-FCMAC). The model is motivated from the famous Chinese ancient Ying-Yang philosophy, which views everything as a product of conflict-harmony process between Ying and Yang. That principle is applied to find the optimal number of clusters and fuzzy sets in the fuzzification phase of the hybrid fuzzy-neural YY-FCMAC network. The analyzed experiment on a set of real traffic data flow of the eastbound Pan Island Expressway (PIE) in Singapore shows the effectiveness of the YY-FCMAC in universal approximation and prediction.