Sequential recommendation methods play an irreplaceable role in recommender systems which can capture the users' dynamic preferences from behavior sequences. Despite their success, these works usually suffer sparsity problem commonly existed real applications. Cross-domain sequential aims to alleviate this by introducing relatively richer source-domain data. However, most existing independently of each domain, may neglect item transition patterns across sequences different domains, i.e., a...

Cross-domain sequential recommendation (CDSR) shifts the modeling of user preferences from flat to stereoscopic by integrating and learning interaction information multiple domains at different granularities (ranging inter-sequence intra-sequence single-domain cross-domain). In this survey, we initially define CDSR problem using a four-dimensional tensor then analyze its multi-type input representations under multidirectional dimensionality reductions. Following that, provide systematic...

Cross-domain sequential recommendation (CDSR) shifts the modeling of user preferences from flat to stereoscopic by integrating and learning interaction information multiple domains at different granularities (ranging inter-sequence intra-sequence single-domain cross-domain). In this survey, we first define CDSR problem using a four-dimensional tensor then analyze its multi-type input representations under multidirectional dimensionality reductions. Following that, provide systematic overview...

Cross-domain sequential recommendation aims to alleviate the sparsity problem while capturing users’ preferences. However, most existing methods learn user preferences in each domain separately, and then perform knowledge transfer between them associate two separated domains, which neglects item transition patterns across sequences from different domains. Moreover, still exists since some items target source domains are interacted with only a limited number of times. To address these issues,...

The problem of single-phase grounding fault arc extinction exists for a long time in the distribution network. To improve effect arc-suppression case occurring network, this paper presents novel comprehensive scheme which can adapt to variation line parameters and transition resistance. When resistance is large, influence voltage drop ignored transfer device used control point zero. If value small, active method will be utilized achieve goal extinguishment. Effectiveness proposed verified by...

In this paper, according to the 'Technical rule for connecting wind farm power system' promulgated by state, least squares support vector machine (LSSVM) algorithm is used build a mathematical model, and six indicators, namely, voltage deviation, fluctuation, flicker, grid harmonics, frequency deviation three-phase imbalance, are comprehensively considered certain quality level calculated, which evaluate of stations substation access points. LSSVM classification method based on principle...

Seismic data is typically displayed as variable density in a three‐dimensional (3‐D) visualization environment. The amplitude of given time instance at seismic trace mapped to gray scale, and section, or the faces volume, becomes two‐dimensional (2‐D) images that are passed into graphics hardware using texture. performs amplitude‐to‐color mapping, primarily bilinear interpolation (which only considers closest 2×2 neighbor data). Unfortunately, introduces noticeable magnification artifacts...

In Meshless natura1 neighbour Petrov-Galerkin method, The natural interpolation is used as trial function anda weak form over the local polygonal sub-domains constructed by Delaunay triangular to obtain discretizedsystem of equilibrium equations, and it’s a new truly meshless method. This method simplified formation theequilibrium facilitates imposition essential boundary conditions system stiffness matrix in thepresent banded sparse. Efforts are made study Pettrov-Galerkin Method,which...

When the active arc suppression device is applied in distribution network, fault characteristics are changed, so existing single line-to-ground line selection methods may misjudge. A novel method based on phase angle of three measurement admittances proposed, which can directly locate section to narrow determination range. The proposed applicable both neutral ineffectively grounded systems and network with application device. judgment result not affected by transition resistance asymmetry...

The traditional single line-to-ground fault phase selection method for distribution network generally ignores the influence of parameter asymmetry. When transition resistance is large enough, it will cause wrong and result in a phase-to-phase short circuit. By measuring neutral point voltage to divide coordinate system, new proposed. proposed has strong adaptability asymmetry three parameters network. can be correctly selected under various grounding modes. Simulations MATLAB/Simulink verify...

The detection and type identification of defects in the external insulation electrical equipment is an important part high-voltage condition assessment. UV pulse technology used to detect analyze discharges by using sensors. In this paper, we propose a method identify polluted insulator discharge based on entropy feature wavelet packet transform. transform decompose reconstruct signal extract three-dimensional time-frequency distribution for analysis, features domain, classify two types...