Neuromorphic computing has come to refer to a variety of brain-inspired computers, devices, and models that contrast the pervasive von Neumann computer architecture This biologically inspired approach has created highly connected synthetic neurons and synapses that can be used to model neuroscience theories as well as solve challenging machine learning problems The promise of the technology is to create a brain-like ability to learn and adapt, but the technical challenges are significant, starting with an accurate neuroscience model of how the brain works, to finding materials and engineering breakthroughs to build devices to support these models, to creating a programming framework so the systems can learn, to creating applications with brain-like capabilities In this work, we provide a comprehensive survey of the research and motivations for neuromorphic computing over its history We begin with a 35-year review of the motivations and drivers of neuromorphic computing, then look at the major research areas of the field, which we define as neuro-inspired models, algorithms and learning approaches, hardware and devices, supporting systems, and finally applications We conclude with a broad discussion on the major research topics that need to be addressed in the coming years to see the promise of neuromorphic computing fulfilled The goals of this work are to provide an exhaustive review of the research conducted in neuromorphic computing since the inception of the term, and to motivate further work by illuminating gaps in the field where new research is needed

A Survey of Neuromorphic Computing and Neural Networks in Hardware.

Performance evaluation of fuzzy classfier systems for multi-dimensional pattern classification problems

We examine the performance of a fuzzy genetics-based machine learning method for multidimensional pattern classification problems with continuous attributes In our method, each fuzzy if-then rule is handled as an individual, and a fitness value is assigned to each rule Thus, our method can be viewed as a classifier system In this paper, we first describe fuzzy if-then rules and fuzzy reasoning for pattern classification problems Then we explain a genetics-based machine learning method that automatically generates fuzzy if-then rules for pattern classification problems from numerical data Because our method uses linguistic values with fixed membership functions as antecedent fuzzy sets, a linguistic interpretation of each fuzzy if-then rule is easily obtained The fixed membership functions also lead to a simple implementation of our method as a computer program The simplicity of implementation and the linguistic interpretation of the generated fuzzy if-then rules are the main characteristic features of our method The performance of our method is evaluated by computer simulations on some well-known test problems While our method involves no tuning mechanism of membership functions, it works very well in comparison with other classification methods such as nonfuzzy machine learning techniques and neural networks

/pdf/performance-evaluation-of-fuzzy-classifier-systems-for-17h2chu5n2.pdf

Performance evaluation of fuzzy classifier systems for multidimensional pattern classification problems

A multiprocessor having an input/output controller, a process controller, and a multidimensional arrays of field programmable gate arrays (FPGAs), each FPGA having its own local memory. The multiprocessor may be programmed to function as a single-instruction, multiple-data (SIMD) parallel processor having a matrix of processing elements (PEs), where each FPGA may be programmed to operate as a submatrix array of PEs. The multiprocessor is especially useful for image processing, pattern recognition, and neural network applications.

FPGA-based processor

Neural networks in designing fuzzy systems for real world applications

An alternative approach for generation of membership functions and fuzzy rules based on radial and cubic basis function networks

An architecture is presented using fuzzy inference methods and neural networks. The combined fuzzy-neural architecture extracts rules from training data and tunes its parameters to obtain optimum results by supervised learning. A prototype system is developed using the proposed architecture, for automatic classification of solder joint images. The classification results obtained are superior to the conventional classifiers and similar to the best results obtained by neural classifiers. This application shows that some of the concerns such as the need for expert knowledge in fuzzy systems and the black box nature in neural networks can be successfully overcome by using fuzzy-neural methods. Additionally, it is possible to include partial a priori knowledge into the network, and to remove superfluous input features from the system, which is a result that cannot be obtained using a conventional neural network. >

A rule based prototype system for automatic classification in industrial quality control

A new method is presented combining the advantages of fuzzy inference and neural network learning. A three-layer radial basis function (RBF) network is used to extract rules and to identify the necessary membership functions of the inputs for a fuzzy classification system. The results obtained applying this new method to IRIS-classification are similar to that of other fuzzy-neural approaches, but only lesser number of rules and membership functions are necessary. This system based on RBF-neurons and extended restricted coulomb energy (RCE) learning allows very fast construction of expert knowledge only from input/output data without externally provided expert help and superfluous input features can be removed automatically after training the network. >

A new method for generating fuzzy classification systems using RBF neurons with extended RCE learning

A new method for generating fuzzy classification systems using rbf neurons with extended rce learning

Methods of hardware implementation for fast, transparent and efficient neural classifiers based on dynamically developing network structures are presented. The neural networks and the learning algorithms are modified for easy hardware implementation. The proposed methods are tested with several application examples. The hardware implementability is verified by simulating the parallel parts in VHDL and generating architectures for field programmable gate arrays (FPGA) with a commercially available high level synthesis tool.

W. Poechmueller

Papers

An alternative approach for generation of membership functions and fuzzy rules based on radial and cubic basis function networks

A rule based prototype system for automatic classification in industrial quality control

A new method for generating fuzzy classification systems using RBF neurons with extended RCE learning

A new method for generating fuzzy classification systems using rbf neurons with extended rce learning

Fuzzy interpretable dynamically developing neural networks with FPGA based implementation