基于多组典型相关变量的因果关系发现算法

现有的因果关系发现算法主要基于单个观察变量本身之间的因果关系,无法适用于多组观察变量,为此提出了一种多组典型相关变量的因果关系发现算法。首先,引入多组典型相关变量建立多组典型相关变量的线性非高斯无环模型并提出对应的目标函数;然后,采用梯度上升的方法求解目标函数,构建多组典型相关变量的因果关系网络。模拟实验验证了该算法的有效性,并在移动基站数据上发现了一批有价值的多组无线网络性能指标间的因果关系。     

一种监督的局部保持典型相关分析算法

从模式分类的角度出发,提出一种监督的局部保持典型相关分析(SLPCCA),通过最大类内成对样本与其近邻间的权重相关性,因而能有效利用样本类别信息的同时保持数据的局部流形结构,并且融合判别型典型相关分析(DCCA)的鉴别信息而不受总类别数的限制。此外,为了提取数据的非线性特征,在核方法的基础上又提出一种核化的SLPCCA(KSLPCCA)。在ORL、Yale、AR和FERET等人脸数据库的实验结果表明,该算法比其他传统的典型相关分析方法具有更好的识别效果。     

Fisher sequential classifiers.

This paper presents further discussion and development of the two-parameter Fisher criterion and describes its two modifications (weighted criterion and another multiclass form). The criteria are applied in two algorithms for training linear sequential classifiers. The main idea of the first algorithm is separating the outermost class from the others. The second algorithm, which is a generalization of the first one, is based on the idea of linear division of classes into two subsets. As linear division of classes is not always satisfactory, a piecewise-linear version of the sequential algorithm is proposed as well. The computational complexity of different algorithms is analyzed. All methods are verified on artificial and real-life data sets.     

Generating combinatorial sets with given properties

Special classes of combinatorial sets called k-sets are analyzed. An algorithm for the generation of k-sets is proposed. It is based on a unified algorithm for generating base combinatorial sets. The possibilities of using it to generate various base sets are considered. The complexity of the algorithms is assessed. The results of computational experiments are analyzed.     

On the generalization of fuzzy rough sets

Rough sets and fuzzy sets have been proved to be powerful mathematical tools to deal with uncertainty, it soon raises a natural question of whether it is possible to connect rough sets and fuzzy sets. The existing generalizations of fuzzy rough sets are all based on special fuzzy relations (fuzzy similarity relations, T-similarity relations), it is advantageous to generalize the fuzzy rough sets by means of arbitrary fuzzy relations and present a general framework for the study of fuzzy rough sets by using both constructive and axiomatic approaches. In this paper, from the viewpoint of constructive approach, we first propose some definitions of upper and lower approximation operators of fuzzy sets by means of arbitrary fuzzy relations and study the relations among them, the connections between special fuzzy relations and upper and lower approximation operators of fuzzy sets are also examined. In axiomatic approach, we characterize different classes of generalized upper and lower approximation operators of fuzzy sets by different sets of axioms. The lattice and topological structures of fuzzy rough sets are also proposed. In order to demonstrate that our proposed generalization of fuzzy rough sets have wider range of applications than the existing fuzzy rough sets, a special lower approximation operator is applied to a fuzzy reasoning system, which coincides with the Mamdani algorithm.     

Cautious transaction schedulers for database concurrency control

Cautious schedulers, which never resort to rollbacks for the purpose of concurrency control, are investigated. In particular, cautious schedulers for classes WW consisting of schedules serializable under the write-write constraints, and WRW, a superclass of W, are considered. The cautious WW-scheduler has a number of nice properties, one of which is the existence of a polynomial-time scheduling algorithm. Since cautious WRW-scheduling is, in general, NP-complete, some restrictions are introduced which allow polynomial-time scheduling. All of these cautious schedulers are based on the assumption that transaction predeclare their read and write sets on arrival. Anomalies which occur when transaction modify their read sets or write sets during execution are discussed and countermeasures are proposed     

面向移动App流量的多特征集合集成聚类方法研究与应用

针对移动互联网流量识别问题,基于多项性能评估指标,分析K-均值和谱聚类算法在不同特征集合或不同识别目标流量数据集上的聚类性能,并提出基于多特征集合的集成聚类方法。比较分析实验表明,相同聚类方法在不同特征集合或不同识别目标数据集上性能有所不同,集成聚类方法能够有效提高利用单个特征集合聚类方法的性能。进一步将集成聚类方法应用于App关联分析,分析结果可为移动App的划分和用户行为分析提供客观依据。     

求极小碰集的蛛网算法

为降低空间复杂度和减少搜索时间,结合极小碰集的特点和生物学中蜘蛛捕食思想,提出了一种搜索极小碰集的蛛网算法。该方法考虑集合之间的相关性,并构造能在蛛网上寻路的访问蜘蛛用于寻找蛛网内集合的所有极小碰集。在该算法中,所提出的访问蜘蛛生成和搜索策略能够降低空间复杂度和减少搜索时间。将此算法与其他的极小碰集算法进行比较,实验结果表明,该算法在保证得到所有极小碰集的前提下,具有较低的空间复杂度和较高的时间效率。     

基于概念格的组规则产生集的算法研究

数据挖掘中传统的关联规则生成算法产生的关联规则集合相当庞大,其中很多规则可由其它规则导出。使用闭项集可以减少规则的数目,而概念格节点间的泛化和例化关系非常适用于规则的提取。目前几种基于概念格的规则提取算法局限于得到准确支持度、信任度的无冗余规则。提出了一种在概念格上挖掘出能推导出所有满足最小支持度、信任度规则的规则产生集算法,文中称之为组规则产生集算法,减少了规则的规模。在此基础上进一步给出了组规则产生集的存储数据结构并用其导出一般规则产生集的算法。     

Exploring constructive cascade networks

Constructive algorithms have proved to be powerful methods for training feedforward neural networks. An important property of these algorithms is generalization. A series of empirical studies were performed to examine the effect of regularization on generalization in constructive cascade algorithms. It was found that the combination of early stopping and regularization resulted in better generalization than the use of early stopping alone. A cubic penalty term that greatly penalizes large weights was shown to be beneficial for generalization in cascade networks. An adaptive method of setting the regularization magnitude in constructive algorithms was introduced and shown to produce generalization results similar to those obtained with a fixed, user-optimized regularization setting. This adaptive method also resulted in the construction of smaller networks for more complex problems. The acasper algorithm, which incorporates the insights obtained from the empirical studies, was shown to have good generalization and network construction properties. This algorithm was compared to the cascade correlation algorithm on the Proben 1 and additional regression data sets.     

Partitioning ideal sets

This paper analyzes a class of approximation algorithms for the NP-complete problem [4] of partitioning a given set of positive real numbers into k subsets. Chandra and Wong [2] analyzed one such algorithm (Graham's LPT rule) for the L ₂ metric on the related classic scheduling problem and proved a 25/24 worst case upper bound in the general case. This result was slightly improved by Leung and Wei [9]. The input sets considered in this paper are sets which have a k-partition with equal sum subsets (termed ideal sets); we prove a new tight upper bound of 37/ 36 for the L ₂ norm on such sets for the entire class of algorithms.     

The Longest Path Problem Is Polynomial on Cocomparability Graphs

The longest path problem is the problem of finding a path of maximum length in a graph. As a generalization of the Hamiltonian path problem, it is NP-complete on general graphs and, in fact, on every class of graphs that the Hamiltonian path problem is NP-complete. Polynomial solutions for the longest path problem have recently been proposed for weighted trees, Ptolemaic graphs, bipartite permutation graphs, interval graphs, and some small classes of graphs. Although the Hamiltonian path problem on cocomparability graphs was proved to be polynomial almost two decades ago, the complexity status of the longest path problem on cocomparability graphs has remained open; actually, the complexity status of the problem has remained open even on the smaller class of permutation graphs. In this paper, we present a polynomial-time algorithm for solving the longest path problem on the class of cocomparability graphs. Our result resolves the open question for the complexity of the problem on such graphs, and since cocomparability graphs form a superclass of both interval and permutation graphs, extends the polynomial solution of the longest path problem on interval graphs and provides polynomial solution to the class of permutation graphs.     

基于标签关联性的分层分类共有与固有特征选择

在大数据时代,数据的样本数量、特征维度和类别数量都在急剧增加,且样本类别间通常存在着层次结构.如何对层次结构数据进行特征选择具有重要意义.近年来,已有相关特征选择算法提出,然而现有算法未充分利用类别的层次结构信息,且忽略了不同类节点具有共有与固有属性的特点.据此,提出了基于标签关联性的分层分类共有与固有特征选择算法.该算法利用递归正则化对层次结构的每个内部节点选择对应的固有特征,并充分利用层次结构分析标签关联性,进而利用正则化惩罚项学习各子树的共有特征.该模型不仅能够处理树结构层次化数据,也能直接处理更为复杂常见的有向无环图结构的层次化数据.在6个树结构数据集和4个有向无环图结构数据集上的实验结果,验证了该算法的有效性.     

Maximum likelihood training of probabilistic neural networks

A maximum likelihood method is presented for training probabilistic neural networks (PNN's) using a Gaussian kernel, or Parzen window. The proposed training algorithm enables general nonlinear discrimination and is a generalization of Fisher's method for linear discrimination. Important features of maximum likelihood training for PNN's are: 1) it economizes the well known Parzen window estimator while preserving feedforward NN architecture, 2) it utilizes class pooling to generalize classes represented by small training sets, 3) it gives smooth discriminant boundaries that often are "piece-wise flat" for statistical robustness, 4) it is very fast computationally compared to backpropagation, and 5) it is numerically stable. The effectiveness of the proposed maximum likelihood training algorithm is assessed using nonparametric statistical methods to define tolerance intervals on PNN classification performance.     

基于矩阵置换的最优无碰撞区跳频序列集的构造

对已有的几类无碰撞区跳频序列集的构造进行推广,提出一种无碰撞区跳频序列集的一般构造。该一般构造是通过对矩阵的列进行置换来实现的。在提出的序列集构造中,序列的长度、序列的条数和无碰撞区大小可灵活变动,而且构造方法多样,序列集的某些性质受具体的构造方法和参数的影响。由该方法得到序列集的参数达到了理论界,是一类最优无碰撞区跳频序列集。     

On the reducibility of sets inside NP to sets with low information content

This paper studies for various natural problems in NP whether they can be reduced to sets with low information content, such as branches, P-selective sets, and membership comparable sets. The problems that are studied include the satisfiability problem, the graph automorphism problem, the undirected graph accessibility problem, the determinant function, and all logspace self-reducible languages. Some of these are complete for complexity classes within NP, but for others an exact complexity theoretic characterization is not known. Reducibility of these problems is studied in a general framework introduced in this paper: prover-verifier protocols with low-complexity provers. It is shown that all these natural problems indeed have such protocols. This fact is used to show, for certain reduction types, that these problems are not reducible to sets with low information content unless their complexity is much less than what it is currently believed to be. The general framework is also used to obtain a new characterization of the complexity class is the class of all logspace self-reducible sets in L^L-sel.     

基于项集优化重组的频繁项集发现算法

发现频繁项集是关联规则挖掘的主要途径,也是关联规则挖掘算法研究的重点。关联规则挖掘的经典Apriori算法及其改进算法大致可以归为基于SQL和基于内存两类。为了提高挖掘效率,在仔细分析了基于内存算法存在效率瓶颈的基础上,提出了一种发现频繁项集的改进算法。该算法使用了一种快速产生和验证候选项集的方法,提高了生成项目集的速度。实验结果显示该算法能有效提高挖掘效率。     

The computation of hitting sets: Review and new algorithms

In model-based diagnosis or other research fields, the hitting sets of a set cluster are usually used. In this paper we introduce some algorithms, including the new BHS-tree and Boolean algebraic algorithms. In the BHS-tree algorithm, a binary-tree is used for the computation of hitting sets, and in the Boolean algebraic algorithm, components are represented by Boolean variables. It runs just for one time to catch the minimal hitting sets. We implemented the algorithms and present empirical results in order to show their superiority over other algorithms for computing hitting sets.     

A comparative analysis of the Global Land Cover 2000 and MODIS land cover data sets

Accurate and up-to-date global land cover data sets are necessary for various global change research studies including climate change, biodiversity conservation, ecosystem assessment, and environmental modeling. In recent years, substantial advancement has been achieved in generating such data products. Yet, we are far from producing geospatially consistent high-quality data at an operational level. We compared the recently available Global Land Cover 2000 (GLC-2000) and MODerate resolution Imaging Spectrometer (MODIS) global land cover data to evaluate the similarities and differences in methodologies and results, and to identify areas of spatial agreement and disagreement. These two global land cover data sets were prepared using different data sources, classification systems, and methodologies, but using the same spatial resolution (i.e., 1 km) satellite data. Our analysis shows a general agreement at the class aggregate level except for savannas/shrublands, and wetlands. The disagreement, however, increases when comparing detailed land cover classes. Similarly, percent agreement between the two data sets was found to be highly variable among biomes. The identified areas of spatial agreement and disagreement will be useful for both data producers and users. Data producers may use the areas of spatial agreement for training area selection and pay special attention to areas of disagreement for further improvement in future land cover characterization and mapping. Users can conveniently use the findings in the areas of agreement, whereas users might need to verify the informaiton in the areas of disagreement with the help of secondary information. Learning from past experience and building on the existing infrastructure (e.g., regional networks), further research is necessary to (1) reduce ambiguity in land cover definitions, (2) increase availability of improved spatial, spectral, radiometric, and geometric resolution satellite data, and (3) develop advanced classification algorithms.     

Sparse CCA using a Lasso with positivity constraints

Canonical correlation analysis (CCA) describes the relationship between two sets of variables by finding linear combinations of the variables with maximal correlation. A sparse version of CCA is proposed that reduces the chance of including unimportant variables in the canonical variates and thus improves their interpretation. A version of the Lasso algorithm incorporating positivity constraints is implemented in tandem with alternating least squares (ALS), to obtain sparse canonical variates. The proposed method is demonstrated on simulation studies and a data set from market basket analysis.