Search

In this paper, we propose a novel method to consistently estimate, at the root- n rate, the coefficient parameters in a biased partial linear single-index model whose error term does not have zero conditional expectation. To achieve this purpose, we first transfer the model to a pro forma linear model and then introduce an artificial variable into a linear bias correction model. Based on the bias correction model, the parameters can then be consistently estimated by the linear least squares method. Both numerical studies and real data analyses are conducted to show the effectiveness of the proposed estimation procedure. [ABSTRACT FROM AUTHOR]

With the growing trend of making vehicles smarter, the idea of utilizing vehicles as the infrastructures for communication and computation has triggered great interest. There have been increasingly efforts integrating the connected vehicles into a cloud computing system, but the performance of such system is restricted by its high latency. To solve this problem, a new computing paradigm, fog computing has been proposed to better exploit potential computing resources of connected vehicles with a collaborative multitude of end-user clients or near-user edge devices [8]. The fog computing differs from cloud computing by its proximity to end users, dense geographical distribution and support for mobility. However, current studies on fog computing based vehicular system mainly focus on its reliability and security issues instead of investigating realistic scenarios. To the best of our knowledge, this paper is the first to propose vehicular fog computing by studying its capabilities using realistic data acquired from tens of thousands of taxis in Beijing, China. A mathematical model is developed for vehicular fog computing, based on which we can make prediction of potential computing capacity of a vehicular fog and analyze the impact of communication range on the capacity. Then we present temporal and spatial distribution of potential computation capacity of vehicular fog computing in a city-wide scale. Our study quantitatively reveals the capabilities of vehicular fog computing at different scales, which offers insightful guidelines for the related system and protocol designs in the future. [ABSTRACT FROM AUTHOR]

• A framework has been developed for extracting the most consistent bits from iris features (IrisCodes). • The proposed model is based on the incorporation of some novel features such as use of iris masks and optimized scoring based on 1D cluster formation. • The results obtained on four benchmark iris databases (CASIAv3 Interval, IITD, MMU2, and CASIAv4 Thousand) demonstrate significant improvements over the baseline EER (%). Iris-based biometric systems are widely considered as one of the most accurate forms for authenticating individual identities. Features from an iris image are commonly represented as a sequence of bits, known as IrisCodes. The work in this paper focuses on locating and subsequently extracting the most consistent bit-locations from these binary iris features. We achieve this objective by initially constructing a Matching-Code vector from some specifically designated training IrisCodes, and subsequently forming a series of 1D clusters in them. Every cluster element is then assigned a score in the range [ 0 − 1 ] on the basis of two cluster properties - the size of the cluster it belongs to and its distance from the center of the cluster. We term this cumulative score as the Significance Index S (b) for a cluster element b. Finally, we select those locations which correspond to the highest scores for every IrisCode. We have tested our approach for four benchmark iris databases (CASIAv3-Interval, CASIAv4-Thousand, IIT Delhi and MMU2) while varying the number of extracted bit-locations from 50 to 300. Our empirical results exhibit significant improvements over the baseline results regarding both the consistency of the extracted bit-locations, as well as the overall performance of the resulting biometric system. [ABSTRACT FROM AUTHOR]

In this paper the comparison of a proposed neural network with generalized type-2 fuzzy weights (NNGT2FW) with respect to the monolithic neural network (NN) and the neural network with interval type-2 fuzzy weights (NNIT2FW) is presented. Generalized type-2 fuzzy inference systems are used to obtain the generalized type-2 fuzzy weights and are designed by a strategy of increasing and decreasing an epsilon variable for obtaining the different sizes of the footprint of uncertainty (FOU) for the generalized membership functions. The proposed method is based on recent approaches that handle weight adaptation using type-1 and type-2 fuzzy logic. The approach is applied to the prediction of the Mackey–Glass time series, and results are shown to outperform the results produced by other neural models. Gaussian noise was applied to the test data of the Mackey–Glass time series for finding out which of the presented methods in this paper shows better performance and tolerance to noise. [ABSTRACT FROM AUTHOR]

Assessing performance of diagnostic markers is a necessary step for their use in decision making regarding various conditions of interest in diagnostic medicine and other fields. Globally useful markers could, however, have ranges of values that are “ diagnostically non-informative” . This paper demonstrates that the presence of marker values from diagnostically non-informative ranges could lead to a loss in statistical efficiency during nonparametric evaluation and shows that grouping non-informative values provides a natural resolution to this problem. These points are theoretically proven and an extensive simulation study is conducted to illustrate the possible benefits of using grouped marker values in a number of practically reasonable scenarios. The results contradict the common conjecture regarding the detrimental effect of grouped marker values during performance assessments. Specifically, contrary to the common assumption that grouped marker values lead to bias, grouping non-informative values does not introduce bias and could substantially reduce sampling variability. The proven concept that grouped marker values could be statistically beneficial without detrimental consequences implies that in practice, tied values do not always require resolution whereas the use of continuous diagnostic results without addressing diagnostically non-informative ranges could be statistically detrimental. Based on these findings, more efficient methods for evaluating diagnostic markers could be developed. [ABSTRACT FROM AUTHOR]

Error correction is typically the first step of de Novo genome assembly from NGS data. This step has an important impact on the quality and speed of the assembly process. However, the majority of available stand-alone error correction solutions can only detect and correct mismatches. Therefore, these solutions only support correcting reads generated by Illumina sequencers. Several solutions support insertions and deletions (indels) and are capable of working with multiple technologies. However, these solutions are limited by correction performance and resource consumption. In this paper, we introduce MuffinEc, an indel-aware multi-technology correction method for NGS data. This method uses a greedy approach to create groups of reads and subsequently corrects them using their consensus. MuffinEc surpasses existing solutions by offering better correction ratios for multiple technologies. This method also exploits parallel processing via OpenMP and uses less computational resources than similar programs, thereby being capable of handling large datasets. MuffinEc is open source and freely available at http://muffinec.sourceforge.net . [ABSTRACT FROM AUTHOR]

In this paper, we propose two novel methods viz., counterpropagation auto-associative neural network (CPAANN) and grey system theory (GST) hybridised with CPAANN for data imputation. The effectiveness of these methods is demonstrated on 12 datasets and the results are compared with that of various extant methods. Wilcoxon signed rank test conducted at 1% level of significance, indicated that the proposed methods are statistically significant against all methods. The spectacular success of CPAANN can be attributed to the local learning, global approximation and auto-association that take place in tandem in a single architecture. Furthermore, significantly CPAANN turned out to be the best in the class of AANN architectures used for imputation. The reason could be the competitive learning that is intrinsic to the CPAANN architecture, but conspicuously absent in other auto-associative neural network architectures. [ABSTRACT FROM AUTHOR]

The least absolute shrinkage and selection operator (LASSO) has been playing an important role in variable selection and dimensionality reduction for linear regression. In this paper we focus on two general LASSO models: Sparse Group LASSO and Fused LASSO, and apply the linearized alternating direction method of multipliers (LADMM for short) to solve them. The LADMM approach is shown to be a very simple and efficient approach to numerically solve these general LASSO models. We compare it with some benchmark approaches on both synthetic and real datasets. [ABSTRACT FROM AUTHOR]

Abstract: Rate variation among the sites of a molecular sequence is commonly found in applications of phylogenetic inference. Several approaches exist to account for this feature but they do not usually enable the investigator to pinpoint the sites that evolve under one or another rate of evolution in a straightforward manner. The focus is on Bayesian phylogenetic mixture models, augmented with allocation variables, as tools for site classification and quantification of classification uncertainty. The method does not rely on prior knowledge of site membership to classes or even the number of classes. Furthermore, it does not require correlated sites to be next to one another in the sequence alignment, unlike some phylogenetic hidden Markov or change-point models. In the approach presented, model selection on the number and type of mixture components is conducted ahead of both model estimation and site classification; the steppingstone sampler (SS) is used to select amongst competing mixture models. Example applications of simulated data and mitochondrial DNA of primates illustrate site classification via ‘augmented’ Bayesian phylogenetic mixtures. In both examples, all mixtures outperform commonly-used models of among-site rate variation and models that do not account for rate heterogeneity. The examples further demonstrate how site classification is readily available from the analysis output. The method is directly relevant to the choice of partitions in Bayesian phylogenetics, and its application may lead to the discovery of structure not otherwise recognised in a molecular sequence alignment. Computational aspects of Bayesian phylogenetic model estimation are discussed, including the use of simple Markov chain Monte Carlo (MCMC) moves that mix efficiently without tempering the chains. The contribution to the field of Bayesian phylogenetics is in (1) the use of mixture models augmented with allocation variables as tools for site classification and quantification of classification uncertainty, (2) the successful application of SS for selection of phylogenetic mixtures, and (3) the development of novel MCMC aspects of relevance to Bayesian phylogenetic models—whether mixtures or not. 1 [1] The MCMC methods discussed in this paper have been coded in a C program; source files are available upon request. Supplementary material is available online (see Appendix A). [Copyright &y& Elsevier]