Linear Combination Of Data Research Articles

Auto-weighted concept factorization for joint feature map and data representation learning

Concept factorization (CF) is an effective matrix factorization model which has been widely used in many applications. In CF, the linear combination of data points serves as the dictionary based on which CF can be performed in both the original feature space as well as the reproducible kernel Hilbert space (RKHS). The conventional CF treats each dimension of the feature vector equally during the data reconstruction process, which might violate the common sense that different features have different discriminative abilities and therefore contribute differently in pattern recognition. In this paper, we introduce an auto-weighting variable into the conventional CF objective function to adaptively learn the corresponding contributions of different features and propose a new model termed Auto-Weighted Concept Factorization (AWCF). In AWCF, on one hand, the feature importance can be quantitatively measured by the auto-weighting variable in which the features with better discriminative abilities are assigned larger weights; on the other hand, we can obtain more efficient data representation to depict its semantic information. The detailed optimization procedure to AWCF objective function is derived whose complexity and convergence are also analyzed. Experiments are conducted on both synthetic and representative benchmark data sets and the clustering results demonstrate the effectiveness of AWCF in comparison with some related models.

IDEG: Integrated Data and Energy Gathering Framework for Practical Wireless Sensor Networks Using Compressive Sensing

Minimizing the energy consumption through efficient data gathering as well as energy harvesting both are open research problems in the context of practical wireless sensor networks (WSNs). Although the number of approaches has been proposed to cater the above, they are limited in terms of computational complexity, energy efficiency, and the ease of implementation. In this paper, we propose an integrated data and energy gathering framework, named hereby as “iDEG,” for practical WSNs. The proposed work is based on compressive sensing and utilizes partial canonical identity (PCI) matrix that samples/picks fewer sensor nodes randomly at any time point instead of picking up a linear combination of data from all sensor nodes at all time points as used in the conventionally popular Gaussian or Bernoulli matrices. This reduces the computation cost, implementation complexity, and energy losses while improving the data recovery. Furthermore, the fewer sensors at any time point selected by PCI contribute for recovering data, while the rest of the nodes harvest energy during those time points. Performance of iDEG has been tested on a real WSN dataset from Intel Lab. Comparative results of iDEG with the conventional approaches highlight its efficacy.

Independent Component Analysis Via Copula Techniques

Independent component analysis (ICA) is a modern factor analysis tool de- veloped in the last two decades. Given p-dimensional data, we search for that linear combination of data which creates (almost) independent components. Here copulae are used to model the p-dimensional data and then independent components are found by optimizing the copula parameters. Based on this idea, we propose the COPICA method for searching independent components. We illustrate this method using several blind source separation examples, which are mathematically equivalent to ICA problems. Finally performances of our method and FastICA are compared to explore the advantages of this method.

Attack on Privacy-Preserving Public Auditing Schemes for Cloud Storage

With the development of Internet, cloud computing has emerged to provide service to data users. But, it is necessary for an auditor on behalf of users to check the integrity of the data stored in the cloud. The cloud server also must ensure the privacy of the data. In a usual public integrity check scheme, the linear combination of data blocks is needed for verification. But, after times of auditing on the same data blocks, based on collected linear combinations, the auditor might derive these blocks. Recently, a number of public auditing schemes with privacy-preserving are proposed. With blinded linear combinations of data blocks, the authors of these schemes believed that the auditor cannot derive any information about the data blocks and claimed that their schemes are provably secure in the random oracle model. In this paper, with detailed security analysis of these schemes, we show that these schemes are vulnerable to an attack from the malicious cloud server who modifies the data blocks and succeeds in forging proof information for data integrity check.

Revision of the Malagasy Camponotus edmondi species group (Hymenoptera, Formicidae, Formicinae): integrating qualitative morphology and multivariate morphometric analysis.

The Malagasy Camponotus edmondi species group is revised based on both qualitative morphological traits and multivariate analysis of continuous morphometric data. To minimize the effect of the scaling properties of diverse traits due to worker caste polymorphism, and to achieve the desired near-linearity of data, morphometric analyses were done only on minor workers. The majority of traits exhibit broken scaling on head size, dividing Camponotus workers into two discrete subcastes, minors and majors. This broken scaling prevents the application of algorithms that uses linear combination of data to the entire dataset, hence only minor workers were analyzed statistically. The elimination of major workers resulted in linearity and the data meet required assumptions. However, morphometric ratios for the subsets of minor and major workers were used in species descriptions and redefinitions. Prior species hypotheses and the goodness of clusters were tested on raw data by confirmatory linear discriminant analysis. Due to the small sample size available for some species, a factor known to reduce statistical reliability, hypotheses generated by exploratory analyses were tested with extreme care and species delimitations were inferred via the combined evidence of both qualitative (morphology and biology) and quantitative data. Altogether, fifteen species are recognized, of which 11 are new to science: Camponotus alamaina sp. n., Camponotus androy sp. n., Camponotus bevohitra sp. n., Camponotus galoko sp. n., Camponotus matsilo sp. n., Camponotus mifaka sp. n., Camponotus orombe sp. n., Camponotus tafo sp. n., Camponotus tratra sp. n., Camponotus varatra sp. n., and Camponotus zavo sp. n. Four species are redescribed: Camponotus echinoploides Forel, Camponotus edmondi André, Camponotus ethicus Forel, and Camponotus robustus Roger. Camponotus edmondi ernesti Forel, syn. n. is synonymized under Camponotus edmondi. This revision also includes an identification key to species for both minor and major castes, information on geographic distribution and biology, taxonomic discussions, and descriptions of intraspecific variation. Traditional taxonomy and multivariate morphometric analysis are independent sources of information which, in combination, allow more precise species delimitation. Moreover, quantitative characters included in identification keys improve accuracy of determination in difficult cases.

BEAM STEERING IN SMART ANTENNAS BY USING LOW COMPLEX ADAPTIVE ALGORITHMS

Array antenna systems are often used to enhance the received signal to interference and noise ratio wh en the signal operates in heavily jammed environment. Proper modeling of the received data at different antenna elements is impo rtant when evaluating the performance of this system, especially when both th e signal and interference have wide frequency bands .. The antenna output is the linear combination of data from all the antenna ele ments. In conventional narrowband beam forming, tim e sequences at different antenna elements are related by some fixed phase sh ift. The phase shift is determined by the wave form s direction of arrival (DOA). In this paper, an efficient method for the pattern synthesis of the linear antenna arrays with the pre scribed nulling and steering lobe is presented. The proposed method is based on Least Me an Square (LMS) algorithm provide a comprehensive atreatment of the signal model used for beam forming, as well as, describing adaptive algorithms to adjust the weigh ts of an array. In order to improve the convergence rate of LMS algorithm in sm art antenna system, in this paper we proposes a ne w normalized LMS (NLMS) algorithm, This new algorithm can be treated as a b lock based simplification of NLMS algorithm which gsatisfactory performance in certain applications in comparison w ith conventional NLMS recursion, i.e., BBNLMS algor ithm. By taking advantage of spatial filtering, the proposed scheme promises to reduce the bandwidth required for transmitting d ata by improving convergence rate. The performance of the BBNLMS algorithm in th e presence of Multi-path effects and multiple users is analyzed using MATLAB simulations. The simulations when compared to that of the LMS algorithm, the results suggest that BB NLMS algorithm can improve the convergence rate and lead to better system effi ciency.

Predicting performance from functional imaging data: methods matter

In the standard approach to functional imaging studies, brain-behavior relationships are studied by contrasting data obtained during different behavioral states. It is generally assumed that relative change yields meaningful data about relevant brain processes, and that the magnitude of the change reflects the extent of a region's involvement in the behavior being studied. The present study takes a different approach by asking the question, Can functional imaging data predict performance? Regional cerebral blood flow was measured using positron emission tomography in a group of 13 right-handed, normal volunteers during speech production and quiet baseline. A number of methodological assumptions were addressed by examining the relationships between different imaging measures derived from the same raw data and performance on the speech task. The results demonstrate that several common assumptions are not necessarily true. First, although measures based on “activated” scans alone had predictive value with respect to speech rate, measures based on contrasts between “baseline” and “activated” states did not. This was true regardless of whether the contrast was based on subtraction or covariance analyses. Second, while many regions demonstrated large signal increases during speech, speech rate could be predicted by a linear combination of data from two regions, neither of which had the highest “activation” peak, and one of which had a negative relationship with performance. The results demonstrate that contrasting experimental conditions do not necessarily isolate or enhance brain activity related to performance, and that the current assumptions about activation in functional imaging need to be reconsidered.

Electronic states and phases of KxC60 from photoemission and X-ray absorption spectroscopy

HIGH-resolution photoemission and soft X-ray absorption spectroscopies have provided valuable information on the electronic structure near the Fermi energy in the superconducting copper oxide compounds1–4, helping to constrain the possible mechanisms of superconductivity. Here we describe the application of these techniques to KxC60, found recently to be superconducting below 19.3 K for x ≈ 3 (refs 5–7). The photoemission and absorption spectra as a function of x can be fitted by a linear combination of data from just three phases, C60, K3C60 and K6C60 indicating that there is phase separation in our samples. The photoemission spectra clearly show a well defined Fermi edge in the K3C60 phase with a density of states of 5.2 x 10-3 electrons eV-1 A-3 and an occupied-band width of 1.2 eV, suggesting that this phase may be a weakly coupled BCS-like (conventional) superconductor. The Cls absorption spectra show large non-rigid-band shifts between the three phases with half and complete filling, in the K3C60 and K6C60 phases respectively, of the conduction band formed from the lowest unoccupied molecular orbital of C60. These observations clearly demonstrate that the conduction band has C 2p character. The non-rigid-band shift coupled with the anomalous occupied-band width implies that there is significant mixing of the electronic states of K and C60 in the superconducting phase.

MESURE DE PROFILS D'HUMIDITÉ ET DE TENSIONS AU COURS D'INFILTRATIONS MONO ET BIDIMENSIONNELLES

Abstract The authors give data of water infiltrations in horizontal and vertical columns, and in a two dimensional model. The soils are two mixtures of sand and loam, and a natural sandy loam. The characteristics of these soils are given in the article. As a conclusion, the two dimensional problem can be approached by a linear combination of data in horizontal and vertical columns.

VI.—On the Graduation of Data by the Orthogonal Polynomials of Least Squares

The problem of fitting a polynomial to data by Least Squares has engaged the attention of many writers. The methods of approach have been many and various. Continued fractions, determinants, the calculus of finite differences and sums, the method of moments, the linear combination of data, the use of the orthogonal polynomials of Legendre and Tchebychef, these and doubtless other instruments of analysis have been pressed into service. At the end of the present paper is given a selective bibliography, which we hope on a future occasion to complete and to supplement by adding brief indications of the standpoint and achievement of each investigator.