Applying machine learning techniques for scaling out data quality algorithms in cloud computing environments

Deduplication is the task of identifying the entities in a data set which refer to the same real world object. Over the last decades, this problem has been largely investigated and many techniques have been proposed to improve the efficiency and effectiveness of the deduplication algorithms. As data sets become larger, such algorithms may generate critical bottlenecks regarding memory usage and execution time. In this context, cloud computing environments have been used for scaling out data quality algorithms. In this paper, we investigate the efficacy of different machine learning techniques for scaling out virtual clusters for the execution of deduplication algorithms under predefined time restrictions. We also propose specific heuristics (Best Performing Allocation, Probabilistic Best Performing Allocation, Tunable Allocation, Adaptive Allocation and Sliced Training Data) which, together with the machine learning techniques, are able to tune the virtual cluster estimations as demands fluctuate over time. The experiments we have carried out using multiple scale data sets have provided many insights regarding the adequacy of the considered machine learning algorithms and proposed heuristics for tackling cloud computing provisioning.     

Finding groups with maximum betweenness centrality

In this paper we consider the problem of identifying the most influential (or central) group of nodes (of some predefined size) in a network. Such a group has the largest value of betweenness centrality or one of its variants, for example, the length-scaled or the bounded-distance betweenness centralities. We demonstrate that this problem can be modelled as a mixed integer program (MIP) that can be solved for reasonably sized network instances using off-the-shelf MIP solvers. We also discuss interesting relations between the group betweenness and the bounded-distance betweenness centrality concepts. In particular, we exploit these relations in an algorithmic scheme to identify approximate solutions for the original problem of identifying the most central group of nodes. Furthermore, we generalize our approach for identification of not only the most central groups of nodes, but also central groups of graph elements that consists of either nodes or edges exclusively, or their combination according to some pre-specified criteria. If necessary, additional cohesiveness properties can also be enforced, for example, the targeted group should form a clique or a κ-club. Finally, we conduct extensive computational experiments with different types of real-life and synthetic network instances to show the effectiveness and flexibility of the proposed framework. Even more importantly, our experiments reveal some interesting insights into the properties of influential groups of graph elements modelled using the maximum betweenness centrality concept or one of its variations.     

Microfluidic device with dual-channel fluorescence acquisition for quantification/identification of cancer cells

This paper presents an optofluidic device for cell discrimination with two independent interrogation regions. Pumping light is coupled to the device, and cell fluorescence is extracted from the two interrogation zones by using optical fibers embedded in the optofluidic chip. To test the reliability of this device, AU-565 cells—expressing EpCAM and HER2 receptors—and RAMOS cells were mixed in a controlled manner, confined inside a hydrodynamic focused flow in the microfluidic chip and detected individually so that they could be discriminated as positive (signal reception from fluorescently labeled antibodies from the AU-565 cells) or negative events (RAMOS cells). A correlation analysis of the two signals reduces the influence of noise on the overall data.     

Increasing selective pressure towards the best compromise in evolutionary multiobjective optimization: The extended NOSGA method

Most current approaches in the evolutionary multiobjective optimization literature concentrate on adapting an evolutionary algorithm to generate an approximation of the Pareto frontier. However, finding this set does not solve the problem. The decision-maker still has to choose the best compromise solution out of that set. Here, we introduce a new characterization of the best compromise solution of a multiobjective optimization problem. By using a relational system of preferences based on a multicriteria decision aid way of thinking, and an outranked-based dominance generalization, we derive some necessary and sufficient conditions which describe satisfactory approximations to the best compromise. Such conditions define a lexicographic minimum of a bi-objective optimization problem, which is a map of the original one. The NOSGA-II method is a NSGA-II inspired efficient way of solving the resulting mapped problem.     

Simultaneous aligning and smoothing of surface triangulations

In this work we develop a procedure to deform a given surface triangulation to obtain its alignment with interior curves. These curves are defined by splines in a parametric space and, subsequently, mapped to the surface triangulation. We have restricted our study to orthogonal mapping, so we require the curves to be included in a patch of the surface that can be orthogonally projected onto a plane (our parametric space). For example, the curves can represent interfaces between different materials or boundary conditions, internal boundaries or feature lines. Another setting in which this procedure can be used is the adaption of a reference mesh to changing curves in the course of an evolutionary process. Specifically, we propose a new method that moves the nodes of the mesh, maintaining its topology, in order to achieve two objectives simultaneously: the piecewise approximation of the curves by edges of the surface triangulation and the optimization of the resulting mesh. We will designate this procedure as projecting/smoothing method and it is based on the smoothing technique that we have introduced for surface triangulations in previous works. The mesh quality improvement is obtained by an iterative process where each free node is moved to a new position that minimizes a certain objective function. The minimization process is done on the parametric plane attending to the surface piece-wise approximation and to an algebraic quality measure (mean ratio) of the set of triangles that are connected to the free node. So, the 3-D local projecting/smoothing problem is reduced to a 2-D optimization problem. Several applications of this method are presented.     

OBSERVER: An Approach for Query Processing in Global Information Systems Based on Interoperation Across Pre-Existing Ontologies

There has been an explosion in the types, availability and volume of data accessible in an information system, thanks to the World Wide Web (the Web) and related inter-networking technologies. In this environment, there is a critical need to replace or complement earlier database integration approaches and current browsing and keyword-based techniques with concept-based approaches. Ontologies are increasingly becoming accepted as an important part of any concept or semantics based solution, and there is increasing realization that any viable solution will need to support multiple ontologies that may be independently developed and managed. In particular, we consider the use of concepts from pre-existing real world domain ontologies for describing the content of the underlying data repositories. The most challenging issue in this approach is that of vocabulary sharing, which involves dealing with the use of different terms or concepts to describe similar information. In this paper, we describe the architecture, design and implementation of the OBSERVER system. Brokering across the domain ontologies is enabled by representing and utilizing interontology relationships such as (but not limited to) synonyms, hyponyms and hypernyms across terms in different ontologies. User queries are rewritten by using these relationships to obtain translations across ontologies. Well established metrics like precision and recall based on the extensions underlying the concepts are used to estimate the loss of information, if any.     

Superpipelined high-performance optical-flow computation architecture

Optical-flow computation is a well-known technique and there are important fields in which the application of this visual modality commands high interest. Nevertheless, most real-world applications require real-time processing, an issue which has only recently been addressed. Most real-time systems described to date use basic models which limit their applicability to generic tasks, especially when fast motion is presented or when subpixel motion resolution is required. Therefore, instead of implementing a complex optical-flow approach, we describe here a very high-frame-rate optical-flow processing system. Recent advances in image sensor technology make it possible nowadays to use high-frame-rate sensors to properly sample fast motion (i.e. as a low-motion scene), which makes a gradient-based approach one of the best options in terms of accuracy and consumption of resources for any real-time implementation. Taking advantage of the regular data flow of this kind of algorithm, our approach implements a novel superpipelined, fully parallelized architecture for optical-flow processing. The system is fully working and is organized into more than 70 pipeline stages, which achieve a data throughput of one pixel per clock cycle. This computing scheme is well suited to FPGA technology and VLSI implementation. The developed customized DSP architecture is capable of processing up to 170 frames per second at a resolution of 800 × 600 pixels. We discuss the advantages of high-frame-rate processing and justify the optical-flow model chosen for the implementation. We analyze this architecture, measure the system resource requirements using FPGA devices and finally evaluate the system’s performance and compare it with other approaches described in the literature.     

Installation d’essais pour la modélisation à échelle réduite des impulsions électromagnétiques

This paper presents a test platform for the scale modeling of electromagnetic pulse effects. The different elements of the installation, i.e. a vertical conical antenna placed on a metallic reference plane, a fast rise time supply generator and a data acquisition and processing facility are described. The field of the antenna has also been studied using a Numerical Electromagnetic Code (Nec)based on the moments method and a comparison with the measurements has been performed. Conclusions resulting from this first stage in the construction of the scale modeling installation are presented.     

Topic model validation

In this paper the problem of performing external validation of the semantic coherence of topic models is considered. The Fowlkes-Mallows index, a known clustering validation metric, is generalized for the case of overlapping partitions and multi-labeled collections, thus making it suitable for validating topic modeling algorithms. In addition, we propose new probabilistic metrics inspired by the concepts of recall and precision. The proposed metrics also have clear probabilistic interpretations and can be applied to validate and compare other soft and overlapping clustering algorithms. The approach is exemplified by using the Reuters-21578 multi-labeled collection to validate LDA models, then using Monte Carlo simulations to show the convergence to the correct results. Additional statistical evidence is provided to better understand the relation of the metrics presented.