Modularity of the Personality Network

The Five Factor Model (FFM) is the most widely used personality model; it proposes a hierarchical structure of personality with personality characteristics, facets, and factors. An increasing number of studies have challenged the FFM and a plethora of factor models with varying numbers of facets and factors have been proposed, leading to uncertainties about the structure of personality. The networked system of interactions between personality characteristics has stimulated promising progresses, however, the methodological developments needed to map the topological structure and functional organization remain scarce. This study aims to explore the hierarchical modular structure of the personality network and the functional role of personality characteristics. A sample of 345,780 individuals ( Mage = 24.99, SDage = 10.00; 59.18% female) that completed the International Personality Item Pool – NEO-120 in a previous study was reanalyzed. A non-regularized method was used to estimate the personality network and ModuLand was used to characterize its modular structure. Results revealed a modular structure comprising three levels: one level with the 120 personality characteristics, a second level with 35 modules, and a third with 9 modules. Such results suggest that specific personality characteristics and modules have specialized roles in the topological structure of the personality network.

A semiparametric model for matrix regression

We focus on regression problems in which the predictors are naturally in the form of matrices. Reduced rank regression and related regularized method have been adapted to matrix regression. However, linear methods are restrictive in their expressive power. In this work, we consider a class of semiparametric additive models based on series estimation of nonlinear functions which interestingly induces a problem of 3rd order tensor regression with transformed predictors. Risk bounds for the estimator are derived and some simulation results are presented to illustrate the performances of the proposed method.

БАГАТОШАРОВА МОДЕЛЬ ТА МЕТОД НАВЧАННЯ ДЛЯ ДЕТЕКТУВАННЯ ШКІДЛИВОГО ТРАФІКУ НА ОСНОВІ АНСАМБЛЮ ДЕРЕВ РІШЕНЬ

The model and training method of multilayer feature extractor and decision rules for a malware traffic detector is proposed. The feature extractor model is based on a convolutional sparse coding network whose sparse encoder is approximated by a regression random forest model according to the principles of knowledge distillation. In this case, an algorithm of growing sparse coding neural gas has been developed for unsupervised training the features extractor with automatic determination of the required number of features on each layer. As for feature extractor, at the training phase to implement of sparse coding the greedy L1-regularized method of Orthogonal Matching Pursuit was used, and at the knowledge distillation phase, the L1-regularized method at the least angles (Least regression algorithm) was additionally used. Due to the explaining-away effect, the extracted features are uncorrelated and robust to noise and adversarial attacks. The proposed feature extractor is unsupervised trained to separate the explanatory factors and allows to use the unlabeled training data, which are usually quite large, with the maximum efficiency. As a model of the decision rules proposed to use the binary encoder of input observations based on an ensemble of decision trees and information-extreme closed hyper-surfaces (containers) for class separation, that are recovery in radial-basis of Hemming' binary space. The addition of coding trees is based on the boosting principle, and the radius of class containers is optimized by direct search. The information-extreme classifier is characterized by low computational complexity and high generalization capacity for small sets of labeled training data. The verification results of the trained model on open CTU test data sets confirm the suitability of the proposed algorithms for practical application since the accuracy of malware traffic detection is 96.1 %.

On a weak solution matching up with the double degenerate parabolic equation

The well-posedness of weak solutions to a double degenerate evolutionary $p(x)$ p ( x ) -Laplacian equation $$ {u_{t}}= \operatorname{div} \bigl(b(x,t){ \bigl\vert {\nabla A(u)} \bigr\vert ^{p(x) - 2}}\nabla A(u)\bigr), $$ u t = div ( b ( x , t ) | ∇ A ( u ) | p ( x ) − 2 ∇ A ( u ) ) , is studied. It is assumed that $b(x,t)| _{(x,t)\in \varOmega \times [0,T]}>0$ b ( x , t ) | ( x , t ) ∈ Ω × [ 0 , T ] > 0 but $b(x,t) | _{(x,t)\in \partial \varOmega \times [0,T]}=0$ b ( x , t ) | ( x , t ) ∈ ∂ Ω × [ 0 , T ] = 0 , $A'(s)=a(s)\geq 0$ A ′ ( s ) = a ( s ) ≥ 0 , and $A(s)$ A ( s ) is a strictly monotone increasing function with $A(0)=0$ A ( 0 ) = 0 . A weak solution matching up with the double degenerate parabolic equation is introduced. The existence of weak solution is proved by a parabolically regularized method. The stability theorem of weak solutions is established independent of the boundary value condition. In particular, the initial value condition is satisfied in a wider generality.

Incremental Multi-Hop Localization Algorithm Based on Regularized Weighted Least Squares

Incremental multi-hop localization algorithm applies to networks with broad range and low density of anchor nodes. However, during the localization process, it tends to be affected by accumulative errors and collinear problem between anchor nodes. We have proposed an incremental multi-hop localization algorithm based on regularized weighted least squares method, and the algorithm uses weighted least squares method to reduce the influence of accumulative errors and uses regularized method to weaken the collinear problem between anchor nodes. The results of both real experiment and simulative experiment show that compared to previous incremental multi-hop localization algorithms, the algorithm proposed in this paper can not only well solve the accumulated errors problem and obtain high localization accuracy, but it has also considered the influence of collinear problem on localization computation during the localization process. We evaluate our method based on various network scenes, and analyze its performance. We also compare our method with several existing methods, and demonstrate the high efficiency of our proposed method.