一种求解动态数据集中k-支配轮廓的算法

在对高维数据集进行轮廓查询时,K-支配轮廓查询算法能够返回较少的轮廓点,有利于用户的决策,但目前的算法都是针对静态数据集设计,无法对动态数据集进行处理.动态数据可分为非数据流数据和数据流数据,本文针对这两种情况提出了相应的增量求解算法,即当数据集发生变化时,以现有的K-支配轮廓为基础,通过对部分数据点进行计算得到新的K-支配轮廓.证明了算法的正确性和有效性,并通过实验对算法进行了分析和验证.     

Approximate probabilistic optimization using exact-capacity-approximate-response-distribution (ECARD)

Probabilistic structural design deals with uncertainties in response (e.g. stresses) and capacity (e.g. failure stresses). The calculation of the structural response is typically expensive (e.g., finite element simulations), while the capacity is usually available from tests. Furthermore, the random variables that influence response and capacity are often disjoint. In previous work we have shown that this disjoint property can be used to reduce the cost of obtaining the probability of failure via Monte Carlo simulations. In this paper we propose to use this property for an approximate probabilistic optimization based on exact capacity and approximate response distributions (ECARD). In Approximate Probabilistic Optimization Using ECARD, the change in response distribution is approximated as the structure is re-designed while the capacity distribution is kept exact, thus significantly reducing the number of expensive response simulations. ECARD may be viewed as an extension of SORA (Sequential Optimization and Reliability Assessment), which proceeds with deterministic optimization iterations. In contrast, ECARD has probabilistic optimization iterations, but in each iteration, the response distribution is approximated so as not to require additional response calculations. The use of inexpensive probabilistic optimization allows easy incorporation of system reliability constraints and optimal allocation of risk between failure modes. The method is demonstrated using a beam problem and a ten-bar truss problem. The former allocates risk between two different failure modes, while the latter allocates risk between members. It is shown that ECARD provides most of the improvement from risk re-allocation that can be obtained from full probabilistic optimization.     

数据流上高效计算子空间Skyline的算法

流数据处理和多维空间中子空间上Skyline的计算是近年来数据管理与数据挖掘领域的研究热点.此前相关工作只专注于滑动窗口上Skyline的维护问题,未涉及到滑动窗口中子空间Skyline的计算.文中提出了一个基于网格索引的高效维护滑动窗口上Skyline的算法,以此为基础采用自顶向下的方式通过两个阶段增量式地返回目标子空间上的结果;开发的多个剪枝策略和启发式优化方法显著地提高了全空间Skyline的维护以及子空间Skyline的计算效率.理论分析和实验结果表明:与同类算法相比,文中提出的StreamSubsky算法以极少的时间开销就能输出第一个结果,并且算法具有良好的可扩展性.     

Approximate Adaptive Output Feedback Stabilization via Passivation of MIMO Uncertain Systems Using Neural Networks

An adaptive output feedback neural network controller is designed, which is capable of rendering affine-in-the-control uncertain multi-input–multi-output nonlinear systems strictly passive with respect to an appropriately defined set. Consequently, a simple output feedback is employed to stabilize the system. The controlled system need not be in normal form or have a well-defined relative degree. Without requiring a zero-state detectability assumption, uniform ultimate boundedness, with respect to an arbitrarily small set, of both the system's state and the output is guaranteed, along with boundedness of all other signals in the closed loop. To effectively avoid possible division by zero, the proposed adaptive controller is of switching type. However, its continuity is guaranteed, thus alleviating drawbacks connected to existence of solutions and chattering phenomena. Simulations illustrate the approach.      

The (Uncertain) Future of Libraries in a Google World

Abstract

Libraries are in competition with other entities that provide information access to students, scholars and the general public. This constitutes a radical change in the information environment. The library profession's failure to take seriously the change that has taken place in the information environment, and to respond with fundamental changes of its own, may well spell disaster for the library as we know it.     

不确定R(o)ssler混沌系统的自适应投影同步

提出了一种不确定R(o)ssler系统自适应投影同步方法.基于Lyapunov稳定性理论,设计了非线性自适应控制器,该控制器可使得驱动系统和具有未知参数的响应系统达到同步,并且可以辨识出响应系统的未知参数.数值模拟验证了所提方案的有效性.     

不确定图数据上的不确定查询处理

图数据结构广泛应用于各种领域的数据建模.由于测量手段和问题特性的限制,数据的不确定性普遍存在.这种不确定性表现在图结构数据中,形成不确定图.之前对于不确定图数据上查询处理的研究,主要是在不确定的图结构数据上查找某一结构确定的图.然而,针对不确定的图数据,其查询很可能也是不确定的.该项工作主要是实现查询过程中的双向匹配,即对于一个不确定的查询,在不确定的图上,得到查询与图的一个可能性最大的匹配组合.这样的研究是具有现实意义的,通过不确定图上对于不确定查询的匹配,可以找到两个不确定结构间存在的最大相似结构,并度量其相似性.     

Approximate Bayes factors for image segmentation: the Pseudolikelihood Information Criterion (PLIC)

We propose a method for choosing the number of colors or true gray levels in an image; this allows fully automatic segmentation of images. Our underlying probability model is a hidden Markov random field. Each number of colors considered is viewed as corresponding to a statistical model for the image, and the resulting models are compared via approximate Bayes factors. The Bayes factors are approximated using BIC (Bayesian Information Criterion), where the required maximized likelihood is approximated by the Qian-Titterington (1991) pseudolikelihood. We call the resulting criterion PLIC (Pseudolikelihood Information Criterion). We also discuss a simpler approximation, MMIC (Marginal Mixture Information Criterion), which is based only on the marginal distribution of pixel values. This turns out to be useful for initialization and it also has moderately good performance by itself when the amount of spatial dependence in an image is low. We apply PLIC and MMIC to a medical image segmentation problem.     

Uncertain reasoning

Annals of Mathematics and Artificial Intelligence -     

Uncertain Distance-Based Range Queries over Uncertain Moving Objects

Distance-based range search is crucial in many real applications. In particular, given a database and a query issuer, a distance-based range search retrieves all the objects in the database whose distances from the query issuer are less than or equal to a given threshold. Often, due to the accuracy of positioning devices, updating protocols or characteristics of applications (for example, location privacy protection), data obtained from real world are imprecise or uncertain. Therefore, existing approaches over exact databases cannot be directly applied to the uncertain scenario. In this paper, we redefine the distance-based range query in the context of uncertain databases, namely the probabilistic uncertain distance-based range (PUDR) queries, which obtain objects with confidence guarantees. We categorize the topological relationships between uncertain objects and uncertain search ranges into six cases and present the probability evaluation in each case. It is verified by experiments that our approach outperform Monte-Carlo method utilized in most existing work in precision and time cost for uniform uncertainty distribution. This approach approximates the probabilities of objects following other practical uncertainty distribution, such as Gaussian distribution with acceptable errors. Since the retrieval of a PUDR query requires accessing all the objects in the databases, which is quite costly, we propose spatial pruning and probabilistic pruning techniques to reduce the search space. Two metrics, false positive rate and false negative rate are introduced to measure the qualities of query results. An extensive empirical study has been conducted to demonstrate the efficiency and effectiveness of our proposed algorithms under various experimental settings.     

Prediction-based Iterative Learning Control (PILC) for Uncertain Dynamic Nonlinear Systems Using System Identification Technique

Prediction-based Iterative Learning Control (PILC) is proposed in this paper for a class of time varying nonlinear uncertain systems. Convergence of PILC is analyzed and the uniform boundedness of tracking error is obtained in the presence of uncertainty and disturbances. It is shown that the learning algorithm not only guarantees the robustness, but also improves the learning rate despite the presence of disturbances and slowly varying desired trajectories in succeeding iterations. The effectiveness of the proposed PILC is presented by simulations.     

Uncertain risk aversion

This paper discusses the risk aversion within the framework of the uncertainty theory (Liu in Uncertainty theory: A branch of mathematics for modeling human uncertainty. Springer, Berlin, 2010b), and introduces the notions of uncertain expected utility and uncertain risk premium. In terms of the Arrow–Pratt index, an uncertain version of Pratt’s theorem is proved, which offers an effective way to make comparisons between different individuals’ risk-averse attitudes. We suggest that uncertain risk aversion can be used to measure human’s risk-averse attitudes when uncertainty exists due to lack of the observed data, just as probabilistic risk aversion when sufficient data can be obtained. Uncertain risk aversion provides an alternative method to compare the risk aversions between individuals under uncertain situations.     

Uncertain Voronoi diagram

In this paper, we introduce the fuzzy Voronoi diagram as an extension of the Voronoi diagram. We assume Voronoi sites to be fuzzy points and then define the Voronoi diagram for this kind of sites, then we provide an algorithm for computing this diagram based on Fortune's algorithm which costs O(nlogn) time. Also we introduce the fuzzy Voronoi diagram for a set of fuzzy circles, rather than fuzzy points, of the same radius. We prove that the boundary of this diagram is formed by the intersection of some hyperbolae, and finally we provide an O(n³logn)-time algorithm to compute the boundary.     

Approximate Musical Evolution

Musical patterns that recur in approximate, rather thanidentical, form within a composition (or body of musical work)are considered to be of considerable importance in musicanalysis. Here we consider the ``evolutionary chain problem':this is the problem of computing a chain of all ``motif'recurrences, each of which is a transformation of (``similar'to) the original motif, but each of which is progressivelyfurther from the original. Here we consider several variants ofthe evolutionary chain problem and we present efficientalgorithms for solving them.     

Approximate boundary-flux calculations

A technique for determining derivatives (fluxes or stresses) from finite element solutions is developed. The approach is a generalization to higher dimensions of a procedure known to give highly accurate results in one dimension. Numerical experiments demonstrate that certain difficulties are associated with corners in the higher-dimensional extensions and two variants of the method are examined. We consider both triangular and quadrilateral elements and observe some interesting differences in the numerical rates of convergence. Finally, this post-processing scheme is tested for nonlinear problems.     

近似多元信息多样性

集成多样性,即个体学习器之间的差异性,是集成学习中的一个基础问题。多元信息多样性(multiinformation diversity)基于信息论来刻画集成多样性,为理解集成多样性提供了一个可行方向,其在实际应用中面临的困难是高阶信息通常难以估计。提出基于一种特殊的k阶t-cherry联结树对高阶信息做低阶近似,从而得到多元信息多样性的近似估计。方法包括基于联结树直接近似估计多元信息和近似估计多元信息分量,并对两者的相关性进行了分析。实验结果表明,在同阶近似下,该估计方法优于现有近似估计方法。     

Approximate Option Pricing

As increasingly large volumes of sophisticated options are traded in world financial markets, determining a ``fair' price for these options has become an important and difficult computational problem. Many valuation codes use the binomial pricing model, in which the stock price is driven by a random walk. In this model, the value of an n -period option on a stock is the expected time-discounted value of the future cash flow on an n -period stock price path. Path-dependent options are particularly difficult to value since the future cash flow depends on the entire stock price path rather than on just the final stock price. Currently such options are approximately priced by Monte Carlo methods with error bounds that hold only with high probability and which are reduced by increasing the number of simulation runs. In this article we show that pricing an arbitrary path-dependent option is \#-P hard. We show that certain types of path-dependent options can be valued exactly in polynomial time. Asian options are path-dependent options that are particularly hard to price, and for these we design deterministic polynomial-time approximate algorithms. We show that the value of a perpetual American put option (which can be computed in constant time) is in many cases a good approximation of the value of an otherwise identical n -period American put option. In contrast to Monte Carlo methods, our algorithms have guaranteed error bounds that are polynomially small (and in some cases exponentially small) in the maturity n . For the error analysis we derive large-deviation results for random walks that may be of independent interest. Received August 13, 1996; revised April 2, 1997.     

Approximate hotlink assignment

Consider a directed rooted tree T=(V,E) of maximal degree d representing a collection V of web pages connected via a set E of links all reachable from a source home page, represented by the root of T. Each leaf web page carries a weight representative of the frequency with which it is visited. By adding hotlinks, shortcuts from a node to one of its descendents, we are interested in minimizing the expected number of steps needed to visit the leaf pages from the home page. We give an O(N²) time algorithm for assigning hotlinks so that the expected number of steps to reach the leaves from the root of the tree is at most H(p)/(log(d+1)−(d/(d+1))logd)+(d+1)/d, where H(p) is the entropy of the probability (frequency) distribution p=〈p₁,p₂,…,p_N〉 on the N leaves of the given tree, i.e., p_i is the weight on the ith leaf. The best known lower bound for this problem is H(p)/log(d+1). We also show how to adapt our algorithm to complete trees of a given degree d and in this case we prove it is optimal, asymptotically in d.     

Approximate Musical Evolution

Musical patterns that recur in approximate, rather thanidentical, form within a composition (or body of musical work)are considered to be of considerable importance in musicanalysis. Here we consider the ``evolutionary chain problem'':this is the problem of computing a chain of all ``motif''recurrences, each of which is a transformation of (``similar''to) the original motif, but each of which is progressivelyfurther from the original. Here we consider several variants ofthe evolutionary chain problem and we present efficientalgorithms for solving them.