Attributed graphs are widely used for the representation of social networks, gene and protein interactions, communication networks, or product co-purchase in web stores. Each object is represented by its relationships...
Attributed graphs are widely used for the representation of social networks, gene and protein interactions, communication networks, or product co-purchase in web stores. Each object is represented by its relationships to other objects (edge structure) and its individual properties (node attributes). For instance, social networks store friendship relations as edges and age, income, and other properties as attributes. these relationships and properties seem to be dependent on each other and exploiting these dependencies is beneficial, e.g. for community detection and community outlier mining. However, state-of-the-art techniques highly rely on this dependency assumption. In particular, community outlier mining is able to detect an outlier node if and only if connected nodes have similar values in all attributes. Such assumptions are generally known as homophily and are widely used. However, looking at multivariate spaces, one can observe that not all given attributes have high dependencies withthe graph structure. For example, social properties such as income or age have strong dependencies withthe graph structure of social networks. In contrast, properties such as gender are rather independent from it. Consequently, recent graph mining algorithms degenerate for multivariate attribute spaces that lack dependency withthe graph structure in some of the attributes.
Vigilance decrement happens in prolonged and monotonous tasks such as driving, therefore efficient estimation of vigilance using machine learning becomes a growing research field in road safety. However, the ground tr...
详细信息
ISBN:
(纸本)9781467372190
Vigilance decrement happens in prolonged and monotonous tasks such as driving, therefore efficient estimation of vigilance using machine learning becomes a growing research field in road safety. However, the ground truth of vigilance level is often unknown. To address the estimation of brain states with unknown ground truth, we proposed an unsupervised manifold clustering method guided by task performance, namely instantaneous lapse rate, without directly using any artificially labels, using electroencephalogram (EEG) as data source. the proposed algorithm utilizes information from bothdata structure and task performance, which is especially suitable for applications with unknown ground truth. Future research directions include using advanced manifold clustering algorithms to increase the robustness towards the high nonlinearity in the EEG feature space and the embedded space, as well as allowing the mapping from multiple clusters to one vigilance level.
the relevance of graphical functions in vehicular applications has increased significantly during the last years. Modern cars are equipped with multiple displays used by different applications such as speedometer, nav...
详细信息
the relevance of graphical functions in vehicular applications has increased significantly during the last years. Modern cars are equipped with multiple displays used by different applications such as speedometer, navigation system, or media players. the rendered output of the applications is stored in so-called off-screen buffers and then bitblitted to the screen buffer at the respective window sizes and positions. To guarantee the visibility of the potentially overlapping windows, the compositing has to match the z-order of the windows. To this end, two common compositing strategies Tile compositing and Full compositing are used, each having performance issues depending on how windows overlap. Since automotive embedded platforms are restricted in power consumption, installation space, and hardware cost, their performance is limited which effectuates the need for highly efficient bitblitting. In order to increase the performance in compositing the windows, we propose Hybrid Compositing which predicts the required bitblitting time and chooses the most efficient strategy for each pair of overlapping windows. Using various scenarios we show that our approach is faster than the other strategies. In addition, we propose CacheHybrid Compositing which reduces the CPU execution time of our approach by up to 66 %. In case of an automotive scenario we show that our optimized approach saves up to 51% bitblitting time compared to existing approaches.
the purpose of this paper is to present knowledge extraction algorithms, dedicated for new electromagnetic system used to evaluate steel bars in reinforced concrete structures. All stages of the rebar identification p...
详细信息
ISBN:
(纸本)9780735412125
the purpose of this paper is to present knowledge extraction algorithms, dedicated for new electromagnetic system used to evaluate steel bars in reinforced concrete structures. All stages of the rebar identification process have been presented. At the first step, relations between parameters of the tested structure and measured waveform are extracted. For this purpose, a dedicated association rules learning algorithm is proposed. In the next stage, the collected data are filtered and smoothed. Finally, classification models are implemented, tested and evaluated. the experimental verification of the applied techniques was carried out, and the selected results are presented.
the U-matrix has become a standard visualization of self-organizing feature maps (SOM). Here we present the abstract U-matrix, which formalizes the structures on a U-matrix such that distance calculations between best...
详细信息
ISBN:
(纸本)9783319076959;9783319076942
the U-matrix has become a standard visualization of self-organizing feature maps (SOM). Here we present the abstract U-matrix, which formalizes the structures on a U-matrix such that distance calculations between best-matching units w.r.t. the height structures of a U-matrix are precisely defined (U-cell distance). this enables the assessment of the topological correctness of the SOM and the implementation of clustering algorithmsthat take the structures seen on the U-matrix into account. A weighted Delaunay graph of the U-cell distances allows the calculation of a dendrogram corresponding to the structures of the U-matrix. the method is shown to detect and visualize meaningful cluster structures on difficult artificial and real-life data.
Progress is presented on the development and implementation of automated data analysis (ADA) software to address the burden in interpreting ultrasonic inspection data for large composite structures. the automated data...
详细信息
ISBN:
(纸本)9780735412125
Progress is presented on the development and implementation of automated data analysis (ADA) software to address the burden in interpreting ultrasonic inspection data for large composite structures. the automated data analysis algorithm is presented in detail, which follows standard procedures for analyzing signals for time-of-flight indications and backwall amplitude dropout. ADA processing results are presented for test specimens that include inserted materials and discontinuities produced under poor manufacturing conditions.
Gene expression data analysis is frequently performed using correlation measures whereas unsupervised and supervised vector quantization methods are usually designed for Euclidean distances. In this paper we summarize...
详细信息
ISBN:
(纸本)9783319076959;9783319076942
Gene expression data analysis is frequently performed using correlation measures whereas unsupervised and supervised vector quantization methods are usually designed for Euclidean distances. In this paper we summarize recent approaches to apply correlation measures to those vector quantization algorithms for analysis of microarray gene expression data. Additionally, we consider k-th order partial correlations as a natural extension if pseudo-correlations should be avoided. Further, we draw the focus to mutual information as powerful alternatives to correlation measures. Related to this we provide the concept of k-th order partial mutual information as counterpart to partial correlations. We apply these methods to an exemplary but real classification problem in gene expression analysis for detection of diabetic patients.
When providing public access to data on the Semantic Web, publishers have various options that include downloadable dumps, Web APIs, and SPARQL endpoints. Each of these methods is most suitable for particular scenario...
When providing public access to data on the Semantic Web, publishers have various options that include downloadable dumps, Web APIs, and SPARQL endpoints. Each of these methods is most suitable for particular scenarios. SPARQL provides the richest access capabilities and is the most suitable option when granular access to the data is needed. However, SPARQL expressivity comes at the expense of high evaluation cost. the potentially large variance in the cost of different SPARQL queries makes guaranteeing consistently good quality of service a very difficult task. Current practices to enhance the reliability of SPARQL endpoints, such as query timeouts and limiting the number of results returned, are far from ideal. they can result in under utilisation of resources by rejecting some queries even when the available resources are sitting idle and they do not isolate "well-behaved" users from "ill-behaved" ones and do not ensure fair sharing among different users. In similar scenarios, where unpredictable contention for resources exists, scheduling algorithms have proven to be effective and to significantly enhance the allocation of resources. To the best of our knowledge, using scheduling algorithms to organise query execution at SPARQL endpoints has not been studied. In this paper, we study, and evaluate through simulation, the applicability of a few algorithms to scheduling queries received at a SPARQL endpoint.
this paper updates my talk on Cache Blocking for Dense Linear algorithms since 1985 given at PPAM 11;see [11]. We again apply Dimension theory to matrices in the Fortran and C programming languages. New data Structure...
详细信息
ISBN:
(数字)9783642552243
ISBN:
(纸本)9783642552243
this paper updates my talk on Cache Blocking for Dense Linear algorithms since 1985 given at PPAM 11;see [11]. We again apply Dimension theory to matrices in the Fortran and C programming languages. New datastructures (NDS) for matrices are given. We use the GCD algorithm to transpose a n by m matrix A in CMO order, standard layout, in-place. Algebra and Geometry are used to make this idea concrete and practical;it is the reason for title of our paper: make a picture of any matrix by the GCD algorithm to convert it into direct sum of square submatrices. the picture is Geometry and the GCD algorithm is Algebra. Also, the in-place transposition of the GKK and TT algorithms will be compared. Finally, the importance of using negative integers will be used to give new results about subtraction and finding primitive roots which also make a priori in-place transpose more efficient.
We live in the era of Big data, or at least our awareness of Big data's presence and impact has sharpened in the past ten years. Compared to data characteristics decades ago, Big data not only means a deluge of un...
详细信息
暂无评论