Improving information-theoretic competitive learning by accentuated information maximization

In this paper, we propose a new computational method for information-theoretic competitive learning. We have so far developed information-theoretic methods for competitive learning in which competitive processes can be simulated by maximizing mutual information between input patterns and competitive units. Though the methods have shown good performance, networks have had difficulty in increasing information content, and learning is very slow to attain reasonably high information. To overcome the shortcoming, we introduce the rth power of competitive unit activations used to accentuate actual competitive unit activations. Because of this accentuation, we call the new computational method “accentuated information maximization”. In this method, intermediate values are pushed toward extreme activation values, and we have a high possibility to maximize information content. We applied our method to a vowel–consonant classification problem in which connection weights obtained by our methods were similar to those obtained by standard competitive learning. The second experiment was to discover some features in a dipole problem. In this problem, we showed that as the parameter r increased, less clear representations could be obtained. For the third experiment of economic data analysis, much clearer representations were obtained by our method, compared with those obtained by the standard competitive learning method.     

A survey on projection neural networks and their applications

Constrained optimization problems arise in numerous scientific and engineering applications, and many papers on the online solution of constrained optimization problems using projection neural networks have been published in the literature. The purpose of this paper is to provide a comprehensive review of the research on projection neural networks for solving various constrained optimizations as well as their applications. Since convergence and stability are important for projection neural networks, theoretical results of projection neural networks are reviewed in detail. In addition, various applications of projection neural networks, e.g., the motion generation of redundant robot manipulators, coordination control of multiple robots with limited communications, generation of winner-take-all strategy, model predictive control and WSN localizations, are discussed and compared. Concluding remarks and future directions of projection neural networks as well as their applications are provided.     

Single-electron winner-take-all network

A winner-take-all (WTA) single-electron neuron is developed for the first time. This new single-electron circuit is proposed in order to implement a WTA neural network with lateral inhibition architecture. An expression for the neuron's activation function is presented. Furthermore, a dot pattern recognition task is successfully performed by the implemented network considering effects such as offset charges and co-tunnelling.     

Are Rule-based Neural Networks Biologically Plausible?

Cognitive models developed in psychology are redefined as mechanisms in artificial intelligence (AI). Brain commands are abstracted in AI as if they were discrete in time and space. To acquire meaning and feel emotions, AI controllers must command a robot. Their symbolic neural networks (SNNs) accumulate invariant temporal rules associating a 'situation' and a 'result' with the 'action' performed. These rules are AI programs learned by experience. An inference engine embodied in these SNNs finds which 'action' produces a given 'intention'. The intentional repetition of brief movements does not require inverse dynamics. Categorization prevents combinatorial explosion. This paper describes a neuroanatomically plausible large-scale architecture integrating SNNs with other NNs. All-or-none irreversible storage is compatible with adaptive learning in this hybrid system. Biological mappings are suggested. Neuromodulators change the processing mode of whole SNNs to enable decisions, freeze states, chain procedure steps and learn temporal rules. Logical impulses are bursts generated by dendritic calcium transients. Synapses transformed are stabilized by self-regulations maintaining multi-stationary states. 'Winner-take-all' sparse coding preserves memory by storing no more than one rule condition per all-or-none synapse.     

Self-organization of Multiple Winner-take-all Neural Networks

In this paper, analysis of the information content of discretely firing neurons in unsupervised neural networks is presented, where information is measured according to the network's ability to reconstruct its input from its output with minimum mean square Euclidean error. It is shown how this type of network can self-organize into multiple winner-take-all subnetworks, each of which tackles only a low-dimensional subspace of the input vector. This is a rudimentary example of a neural network that effectively subdivides a task into manageable subtasks.     

一种高性能CMOS电流模Winner-take-all电路

提出了一种新颖的CMOS电流模Winner-take-all (WTA) 电路．该电路利用可再生比较器提高了电路的解析度和速度,没有使用电流镜,而是利用整流电路输出电流,从而改善了电路的精确度,不同于传统的树型WTA电路．还提出了一种新的并行N输入WTA结构．在TSMC 0.18μm CMOS工艺下设计了一个8输入WTA电路,并与已有的WTA电路进行了比较．仿真结果表明,该电路可以达到1nA的解析度和99.99%的精确度,同时面积小,功耗低,非常适合于各种嵌入式智能应用．     

基于软竞争机制的对传网络模型及应用

针对传统对传网络(CounterPropagationNetwork,CPN)模型和学习算法存在的问题和不足,提出改进模型及竞争层的改进算法。在竞争层使用软竞争机制得到竞争层的输出,克服传统CPN使用胜者全得竞争机制的弊病,使竞争层中每一个神经元节点能充分发挥作用,参与网络的训练和权值的调整,提高竞争层中神经元的利用率,使网络能实现运用最少的神经元,达到要求的性能。从数值实验的对比看出,由于改进了网络模型和竞争算法,增强了CPN的模拟精度,CPN能更好地逼近模拟函数,提高了CPN的使用效率,网络性能得到了很大的提高。     

基于重排序Census变换的半全局立体匹配算法

摘　要针对立体匹配中传统局部算法在计算匹配代价时精度低、抗噪能力弱等问题,提出一种结合改进的Census变换和单方向动态规划优化的半全局立体匹配算法。首先,重排序不同尺度的Census变换窗口中的像素,取其中值计算Hamming距,解决了传统算法对Census变换窗口中心像素依赖的问题。其次,基于单方向动态规划的路径聚合算法对初始代价值进行优化,减少初始代价值中的异常匹配点,提高对弱纹理部分的视差重建,进一步提高匹配精度。最后,采用赢者通吃策略选择单个像素最小代价聚合值所对应的视差,并在视差优化阶段基于左右一致性原则剔除错误视差。实验结果表明,改进的半全局立体匹配算法生成的初始视差图平均误匹配率降低了8.22%,质量相对更高;且在不同噪声下的平局误匹配率均在8%以下,有效的增强了抗噪声的鲁棒性,提升了匹配精度。