Uncertain distance-based outlier detection with arbitrarily shaped data objects

Enabling information systems to face anomalies in the presence of uncertainty is a compelling and challenging task. In this work the problem of unsupervised outlier detection in large collections of data objects modeled by means of arbitrary multidimensional probability density functions is considered. We present a novel definition of uncertain distance-based outlier under the attribute level uncertainty model, according to which an uncertain object is an object that always exists but its actual value is modeled by a multivariate pdf. According to this definition an uncertain object is declared to be an outlier on the basis of the expected number of its neighbors in the dataset. To the best of our knowledge this is the first work that considers the unsupervised outlier detection problem on data objects modeled by means of arbitrarily shaped multidimensional distribution functions. We present the UDBOD algorithm which efficiently detects the outliers in an input uncertain dataset by taking advantages of three optimized phases, that are parameter estimation, candidate selection, and the candidate filtering. An experimental campaign is presented, including a sensitivity analysis, a study of the effectiveness of the technique, a comparison with related algorithms, also in presence of high dimensional data, and a discussion about the behavior of our technique in real case scenarios.

     

Connectivity compression in an arbitrary dimension

This paper presents a general lossless connectivity compression scheme for manifolds in any dimension with arbitrary cells, orientable or not, with or without borders. Relying on a generic topological model called generalized maps, our method performs a region-growing traversal of its primitive elements while describing connectivity relations with symbols. The set of produced symbols is compressed using standard data compression techniques. These algorithms have been successfully applied to various models (surface, tetrahedral and hexahedral meshes), showing the efficiency and genericity of the proposed scheme.     

Real-time rendering of flames on arbitrary deformable objects

This paper proposes a new real-time algorithm to generate visually plausible flames on arbitrary deformable objects.In order to avoid the time-consuming computation of physical fields,the main idea of our algorithm is to build an approximate distance field based on the object’s surface.And then the distance field is sampled and the distance samples are employed to fetch values from a color map which is precomputed according to physical methods.In order to simulate the dynamic flames by static distance field,simplex noise is used to disturb the sampling process.Our algorithm is also capable of handling the interaction between the flames and external factors such as wind.In order to achieve such a goal,two approximate distance fields are built to represent the inner flames and the outer flames respectively,which are combined together to accomplish the interaction.The experimental results show that our algorithm can produce visually plausible and user controllable flames on arbitrary deformable objects in real-time.     

Fast and accurate approximation of digital shape thickness distribution in arbitrary dimension

We present a fast and accurate approximation of the Euclidean thickness distribution computation of a binary shape in arbitrary dimension. Thickness functions associate a value representing the local thickness for each point of a binary shape. When considering with the Euclidean metric, a simple definition is to associate with each point x, the radius of the largest ball inscribed in the shape containing x. Such thickness distributions are widely used in many applications such as medical imaging or material sciences and direct implementations could be time consuming. In this paper, we focus on fast algorithms to extract such distribution on shapes in arbitrary dimension.     

Simulations of dynamically triangulated gravity — an algorithm for arbitrary dimension

Recent models for discrete Euclidean quantum gravity incorporate a sum over simplicial triangulations. We describe an algorithm for simulating such models in arbitrary dimension. As illustration we show results from simulations in four dimensions.     

On approximating Euclidean metrics by weighted t-cost distances in arbitrary dimension

In this work, we propose a new class of distance functions called weighted t-cost distances. This function maximizes the weighted contribution of different t-cost norms in n-dimensional space. With proper weight assignment, this class of function also generalizes m-neighbor and octagonal distances. A non-strict upper bound (denoted as R_u in this work) of its relative error with respect to Euclidean norm is derived and an optimal weight assignment by minimizing R_u is obtained. However, it is observed that the strict upper bound of weighted t-cost norm may be significantly lower than R_u. For example, an inverse square root weight assignment leads to a good approximation of Euclidean norm in arbitrary dimension.     

An improved midpoint rasterization algorithm implemented in FPGA

With the continuous progress of technology and the improvement of people's consumption level, requirements for picture quality are constantly increasing, which has raised the requirements for graphic processing units(GPU) and promoted their development. Among functional modules of GPU, the rasterization module is an important part of GPU. In this paper, we first review the development of rasterization algorithms and analyze their advantages and disadvantages. After that, we improve classical midpoint traversal by adding the idea of tile scanning and improving the parallelism of the algorithm and propose a tile-based multi-parallel midpoint traversal. Then, we implement and verify the proposed algorithm on software and hardware and compare the performance of the proposed algorithm with other algorithms to verify the superiority of the proposed algorithm. Finally, we deploy the improved algorithm in FPGA to verify the feasibility of the hardware.     

改进的基于距离的关联规则聚类

关联规则挖掘会产生大量的规则,为了从这些规则中识别出有用的信息,需要对规则进行有效的分类组织.现有的规则聚类方法往往直接计算规则间的距离,忽略了项与项之间的联系,不能精确得出规则间的距离.提出一种改进的规则间距离的度量方法,首先计算项间的距离,其次计算相集间的距离和规则间的距离,最后基于此距离利用DBSCAN算法对关联规则进行聚类.实验结果表明,此方法是有效可行的,并能准确发现孤立规则.     

Continuous distance-based skyline queries in road networks

In recent years, the research community has introduced various methods for processing skyline queries in road networks. A skyline query retrieves the skyline points that are not dominated by others in terms of static and dynamic attributes (i.e., the road distance). This paper addresses the issue of efficiently processing continuous skyline queries in road networks. Two novel and important distance-based skyline queries are presented, namely, the continuous  _dε-skyline$d_{\epsilon }\mathrm{-}\mathit{skyline}$query   (_Cdε-SQ${\mathit{Cd}}_{\epsilon }\mathrm{-}\mathit{SQ}$) and the continuous k nearest neighbor-skyline query (Cknn-SQ  ). A grid index is first designed to effectively manage the information of data objects and then two algorithms are proposed, the _Cdε-SQ${\mathit{Cd}}_{\epsilon }\mathrm{-}\mathit{SQ}$algorithm   and the _Cdε-SQ⁺${\mathit{Cd}}_{\epsilon }\mathrm{-}{\mathit{SQ}}^{+}$algorithm  , which are combined with the grid index to answer the _Cdε-SQ${\mathit{Cd}}_{\epsilon }\mathrm{-}\mathit{SQ}$. Similarly, the Cknn-SQ algorithm and the Cknn-SQ⁺algorithm are developed to efficiently process the Cknn-SQ. Extensive experiments using real road network datasets demonstrate the effectiveness and the efficiency of the proposed algorithms.     

Fast mining of distance-based outliers in high-dimensional datasets

Defining outliers by their distance to neighboring data points has been shown to be an effective non-parametric approach to outlier detection. In recent years, many research efforts have looked at developing fast distance-based outlier detection algorithms. Several of the existing distance-based outlier detection algorithms report log-linear time performance as a function of the number of data points on many real low-dimensional datasets. However, these algorithms are unable to deliver the same level of performance on high-dimensional datasets, since their scaling behavior is exponential in the number of dimensions. In this paper, we present RBRP, a fast algorithm for mining distance-based outliers, particularly targeted at high-dimensional datasets. RBRP scales log-linearly as a function of the number of data points and linearly as a function of the number of dimensions. Our empirical evaluation demonstrates that we outperform the state-of-the-art algorithm, often by an order of magnitude.     

A distance-based approach to entity reconciliation in heterogeneousdatabases

In modern organizations, decision makers must often be able to quickly access information from diverse sources in order to make timely decisions. A critical problem facing many such organizations is the inability to easily reconcile the information contained in heterogeneous data sources. To overcome this limitation, an organization must resolve several types of heterogeneity problems that may exist across different sources. We examine one such problem called the entity heterogeneity problem, which arises when the same real-world entity type is represented using different identifiers in different applications. A decision-theoretic model to resolve the problem is proposed. Our model uses a distance measure to express the similarity between two entity instances. We have implemented the model and tested it on real-world data. The results indicate that the model performs quite well in terms of its ability to predict whether two entity instances should be matched or not. The model is shown to be computationally efficient. It also scales well to large relations from the perspective of the accuracy of prediction. Overall, the test results imply that this is certainly a viable approach in practical situations     

基于UML & SystemC的GPU几何管线光栅化硬件建模

架构与算法是决定GPU性能的重要因素,需要尽可能早地对其进行评估和验证。提出基于统一建模语言(Unified Modeling Language, UML)的模型,详述了针对GPU几何管线架构和线图元光栅化算法建模的过程及方法,并采用SystemC语言实现了事务级建模(Transaction-level Modeling,TLM)模型和仿真。验证了架构和算法的正确性以及模型的有效性和可行性,为RTL设计提供了参考依据。     

Optimal separable algorithms to compute the reverse euclidean distance transformation and discrete medial axis in arbitrary dimension

In binary images, the distance transformation (DT) and the geometrical skeleton extraction are classic tools for shape analysis. In this paper, we present time optimal algorithms to solve the reverse Euclidean distance transformation and the reversible medial axis extraction problems for d-dimensional images. We also present a d-dimensional medial axis filtering process that allows us to control the quality of the reconstructed shape     

基于融合学习的无监督多维时间序列异常检测

随着多云时代的到来,云际智能运维能够提前检测处理云平台的故障,从而确保其高可用性. 由于云系统的复杂性,运维数据在数据局部性和数据全局性上呈现出多样的时间依赖和维度间依赖,这给多维时间序列异常检测带来很大的挑战. 然而,现有的多维时间序列异常检测方法大多是从正常时序数据中学习到特征表示并基于重构误差或预测误差检测异常,这些方法无法同时捕获多维时间序列在局部性和全局性上的信息依赖,从而导致异常检测效果差. 针对上述问题,提出了一种基于融合学习的无监督多维时间序列异常检测方法,同时对多维时间序列的数据局部特征和数据全局特征进行建模,得到更加丰富的时序重构信息,并基于重构误差检测异常. 具体地,通过在时域卷积网络中引入自注意力机制使得模型在构建局部关联性的同时更加关注数据全局特征,并在时域卷积模块和自注意力模块间加入信息共享机制实现信息融合,从而能够更好地对多维时序的正常模式进行重构. 在多个多维时间序列真实数据集上的实验结果表明,相较于之前的多维时间序列异常检测,提出的方法在F1分数上提升了高达0.0882.

     

Tree-based versus distance-based key recognition in musical audio

A tree-based method for the recognition of the tonal center or key in a musical audio signal is presented. Time-varying key feature vectors of 264 synthesized sounds are extracted from an auditory-based pitch model and converted into character strings using PCA-analysis and classification trees. The results are compared with distance-based methods. The characteristics of the new tonality analysis tool are illustrated on various examples. The potential of this method as a building stone in a music retrieval system is discussed.     

An intuitive distance-based explanation of opposition-based sampling

The impact of the opposition concept can be observed in many areas around us. This concept has sometimes been called by different names, such as, opposite particles in physics, complement of an event in probability, absolute or relative complement in set theory, and theses and antitheses in dialectic. Recently, opposition-based learning (OBL) was proposed and has been utilized in different soft computing areas. The main idea behind OBL is the simultaneous consideration of a candidate and its corresponding opposite candidate in order to achieve a better approximation for the current solution. OBL has been employed to introduce opposition-based optimization, opposition-based reinforcement learning, and opposition-based neural networks, as some examples among others. This work proposes an Euclidean distance-to-optimal solution proof that shows intuitively why considering the opposite of a candidate solution is more beneficial than another random solution. The proposed intuitive view is generalized to N-dimensional search spaces for black-box problems.     

Efficient distance-based representative skyline computation in 2D space

Representative skyline computation is a fundamental issue in database area, which has attracted much attention in recent years. A notable definition of representative skyline is the distance-based representative skyline (DBRS). Given an integer k, a DBRS includes k representative skyline points that aims at minimizing the maximal distance between a non-representative skyline point and its nearest representative. In the 2D space, the state-of-the-art algorithm to compute the DBRS is based on dynamic programming (DP) which takes O(k m ²) time complexity, where m is the number of skyline points. Clearly, such a DP-based algorithm cannot be used for handling large scale datasets due to the quadratic time cost. To overcome this problem, in this paper, we propose a new approximate algorithm called ARS, and a new exact algorithm named PSRS, based on a carefully-designed parametric search technique. We show that the ARS algorithm can guarantee a solution that is at most ?? larger than the optimal solution. The proposed ARS and PSRS algorithms run in O(klog2mlog(T/??)) and O(k ² log3m) time respectively, where T is no more than the maximal distance between any two skyline points. We also propose an improved exact algorithm, called PSRS+, based on an effective lower and upper bounding technique. We conduct extensive experimental studies over both synthetic and real-world datasets, and the results demonstrate the efficiency and effectiveness of the proposed algorithms.     

Emphasizing the rank positions in a distance-based aggregation procedure

This paper deals with the problem of aggregating individual preferences in order to obtain a social order. In particular, a preference aggregation procedure is proposed for those cases in which the decision-makers express their preferences by means of a ranking of alternatives. Among the most commonly applied methods for this purpose are those based on distance measures between individual and collective preferences, which look for the solution that minimizes the disagreement across decision-makers. This class of procedures may include weighting factors in order to emphasize the relative importance of the individuals. In the model proposed here, a weighted disagreement function that computes the differences between alternatives differentiating the rank positions of the alternatives is developed. The proposed disagreement function weighs the differences between orders depending on the ordinal position that the alternative occupies.