Probabilistic neural computing with advanced nanoscale MOSFETs

The use of intrinsic nanoscale MOSFET noise for probabilistic computation is explored, using the continuous restricted Boltzmann machine (CRBM), a probabilistic neural model, as the exemplar architecture. The CRBM is modified by localising noise in its synaptic multipliers, exploiting random telegraph signal (RTS) noise in nanoscale MOSFETs. A look-up table (LUT) technique is adopted to link temporal noise data to the synaptic multipliers of a CRBM, trained to model simple, non-trivial data distributions. It is shown that, for such distributions at least, the CRBM with intrinsic nanoscale MOSFET noise can be trained to provide a useful model.     

The pRAM: an adaptive VLSI chip

The pRAM (probabilistic RAM) is a nonlinear stochastic device with neuron like behavior. The pRAM is realizable in hardware, and the third-generation VLSI pRAM chip is described. This chip is adaptive since learning algorithms have been incorporated on-chip, using reinforcement training. The pRAM chip is also adaptive with respect to the interconnections between neurons. Results achieved from a small net of pRAM's performing a pattern-recognition task using reinforcement training are presented.     

基于一种连续自编码网络的图像降维和重构

针对高维连续数据的降维问题,提出一种新的非线性降维方法,称为连续自编码(Continuous autoencoder,C-autoencoder)神经网络,该方法采用限制玻耳兹曼机的连续形式(Continuous restricted Boltzmann machine,CRBM)的网络结构,通过训练具有多个中间层的双向深层神经网络将高维连续数据转换成低维嵌套并继而重构高维连续数据.这种连续自编码网络提供了高维连续数据空间和低维嵌套结构的双向映射,有效解决了大多数非线性降维方法所不具备的逆向映射问题,特别适用于高维连续数据的降维和重构.将C-autoencoder用于连续帧图像的实验表明,C-autoencoder不仅能发现嵌入在高维连续帧图像中的非线性低维结构,也能有效地从低维结构中恢复原始高维图像数据,而且还能对连续帧图像有效地进行内插重构.     

On the Handling of Continuous-Valued Attributes in Decision Tree Generation

We present a result applicable to classification learning algorithms that generate decision trees or rules using the information entropy minimization heuristic for discretizing continuous-valued attributes. The result serves to give a better understanding of the entropy measure, to point out that the behavior of the information entropy heuristic possesses desirable properties that justify its usage in a formal sense, and to improve the efficiency of evaluating continuous-valued attributes for cut value selection. Along with the formal proof, we present empirical results that demonstrate the theoretically expected reduction in evaluation effort for training data sets from real-world domains.     

Convolutional restricted Boltzmann machines learning for robust visual tracking

It is a critical step to choose visual features in object tracking. Most existing tracking approaches adopt handcrafted features, which greatly depend on people’s prior knowledge and easily become invalid in other conditions where the scene structures are different. On the contrary, we learn informative and discriminative features from image data of tracking scenes itself. Local receptive filters and weight sharing make the convolutional restricted Boltzmann machines (CRBM) suit for natural images. The CRBM is applied to model the distribution of image patches sampled from the first frame which shares same properties with other frames. Each hidden variable corresponding to one local filter can be viewed as a feature detector. Local connections to hidden variables and max-pooling strategy make the extracted features invariant to shifts and distortions. A simple naive Bayes classifier is used to separate object from background in feature space. We demonstrate the effectiveness and robustness of our tracking method in several challenging video sequences. Experimental results show that features automatically learned by CRBM are effective for object tracking.     

VLSI数据库系统的物理设计和实现技术

在VLSI芯片的设计过程中,所牵涉到的设计数据的数量是巨大的,数据库管理系统则起着十分重要和不可替代的作用。本文依据VLSI设计过程的特点和要求以及设计数据的类型与特征,研讨适用于VLSI CAD系统中的数据库管理系统的物理设计和实现问题,同时给出一个实际设计的VLSI数据库系统的数据模型和组织方式以及物理结构。     

A knowledge-based expert system for automatic visual VLSI reverse-engineering: VLSI layout version

This paper presents a method of knowledge representation for very large scale integration (VLSI) chip design which provides the necessary information for abstraction from the physical design to gate-level logic through a high-level behavioral model. The representation scheme used by the ANTISTROFEAS system utilizes a hierarchical attributed graph structure which consists of incrementally abstracted design information for the VLSI system. This method of knowledge representation is well-suited to reverse-engineering of VLSI chips from the layer mask layout data, but is also applicable to applications at many levels of the design process including design rule checking, logic synthesis, design verification, and partitioning-compaction problems. The representation scheme is applicable to any VLSI technology, and is designed to take advantage of artificial intelligence. expert system techniques, by disassociating the representation and manipulation of the VLSI design data from the rules which govern its correctness and transformation for other usage.     

Congestion estimation for hexagonal routing

Interconnect planning is an important issue in VLSI designs. Routing congestion is an important problem during placement. This paper presents a novel approach to estimating the routing congestion on the hexagonal model with bounded-length detours. It is the first probabilistic estimation work on hexagonal topologies. A combinatorial estimation algorithm is presented. The experimental results demonstrate the effectiveness of the method.     

一种基于云模式连续型属性离散化的算法

在数据挖掘研究过程中，对连续型属性一般要进行离散化。特别是在模糊数据挖掘中，还要对离散化的区间进行模糊处理。文中依托云模式，并结合粗糙集理论提出一种新的连续型属性离散化算法。     

SAFARI: a structured approach for automatic rule

This paper describes a new algorithm for obtaining rules automatically from training examples. The algorithm is applicable to examples involving both objects: with discrete and continuous-valued attributes. The paper explains a new quantization procedure fur continuous-valued attributes and shows how appropriate ranges of values of various attributes are obtained. The algorithm uses a decision-tree-based approach for obtaining rules, but unlike other tree-based algorithms such as ID3, it allows more than one attribute at a node which greatly improves its performance. The ability of the algorithm to obtain a measure of partial match further enhances its generalization characteristic. The algorithm produces the same rules irrespective of the order of presentation of training examples. The algorithm has been demonstrated on classification problems. The results have compared favorably with those obtained by existing inductive learning algorithms.     

Dynamic probabilistic networks for modelling and identifying dynamic systems: a MCMC approach

In this article we deal with the problem of interpreting data coming from a dynamic system by using causal probabilistic (CPN), a probabilistic graphical model particularly appealing in Intelligent Data Analysis. We discuss the different approaches presented in the literature, outlining their pros and cons through a simple training example. Then, we present a new method for reconstructing the state of the dynamic system, based on Markov Chain Monte Carlo algorithms, called dynamic probabilistic network smoothing (DPN-smoothing). Finally, we present an example of the application of DPN-smoothing in the field of signal deconvolution.     

一种基于哈希链表的高效概念漂移连续属性处理算法

本文重点研究了数据流挖掘中存在概念漂移情形的连续属性处理算法。数据流是一种增量、在线、实时的数据模型。VFDT是数据流挖掘中数据呈稳态分布情形下最成功的算法之一;CVFDT是有效解决数据流挖掘中概念漂移问题的算法之一。基于CVFDT,本文提出了有效地解决数据流挖掘中存在概念漂移情形的连续属性处理问题的扩展哈希表算法HashCVFDT。该算法在属性值插入、查找和删除时具有哈希表的快速性,而在选取每个连续属性的最优化划分节点时解决了哈希表不能有序输出的缺点。     

HDL描述中决策图模型的提取方法和实现

决策图模型描述了TVLSI设计信号间的数据依赖关系，在VLSI设计验证中有广泛的应用。该文扩展了已有的决策图模型，给出了从硬件描述语言描述的设计中抽取决策图模型的方法并加以实现。将该方法用于RTL级模拟矢量自动生成系统，取得了良好的效果。     

基于卷积受限玻尔兹曼机的医学图像分类新方法

利用数据挖掘方法对医学图像做分析是目前研究的热点之一,常用的挖掘方法首先需要从医学图像中提取特征,然后进行分类分析。目前,应用最多的是提取图像的统计特征,这种方法对所提取的特征有很强的依赖性。采用一种深度学习的新方法——卷积受限玻尔兹曼机模型,并且采用改进的快速持续对比散度算法对模型进行训练。该方法直接从乳腺X光图像中自主学习特征并利用学习到的特征对图像进行分类。实验结果显示,新方法对医学图像的分类精度相对于已有方法有明显的提升。     

基于概率神经网络的入侵检测技术

提出一种基于概率神经网络的高效入侵检测技术。对网络数据处理、概率神经网络的训练与检测及其算法进行分析。在网络训练中,提出一种基于实验数据选择概率神经网络关键参数的方法,分析该方法的可行性。实验表明通过此方法能使入侵检测系统具有更高的检测精度和效率。     

适用于VLSI设计的LVV数据模型及其数据库管理系统

本文分析了VLSI设计中的数据类型和相互关系,提出了适用于VLSI设计的LVV数据模型,它包含对象、版本、视图和文档四个基本概念,支持面向对象的数据操作,描述设计对象的层次结构和设计衍变过程,且根据模型的语义性可进行数据完整性及描述等价性的控制.LVV系统是建立在LVV模型基础上的数据库管理系统,除上述数据模型所提供的特点外,还具有统一的用户界面和数据共享性好等特点.     

连续值Luk系统中公式的Γ-真度及其性质

对Lukasiewicz系统中的真度理论进行了进一步的研究。将n值Lukasiewicz系统中公式相对于有限理论的ΣΓ-真度理论与连续值Lukasiewicz系统中公式的积分真度相结合,在连续值Lukasiewicz系统中引入了公式相对于有限理论的Γ-真度理论,讨论了其中的主要性质;在公式集F(S)上引入了任意两公式相对于有限理论的Γ-伪距离,从而拓宽了真度理论的应用范围。     

Stability evaluation and track life of the PDAF for tracking inclutter

The authors present an effective approach of a hybrid nature to the nonsimulation performance evaluation of the probabilistic data association filtering (PDAF) method for tracking in clutter. In this approach, a continuous-valued covariance, which is a function of a discrete-valued random variable (the number of validated measurements), is used to characterize the tracking errors in an average sense. This covariance can be calculated offline recursively from a modified Riccati equation, which can be obtained by replacing the measurement-dependent terms in the original stochastic equation with their conditional expectations. This approach has the merit that it yields a quantification of the transients of tracking divergence as well as substantially better accuracy than previous work. Such an approach is particularly suitable for stability evaluation of tracking filters. In addition, a quantitative study of the track-life problem is made in which the number of validated measurements plays a central role