Vehicle Interaction Behavior Prediction with Self-Attention

The structured road is a scene with high interaction between vehicles, but due to the high uncertainty of behavior, the prediction of vehicle interaction behavior is still a challenge. This prediction is significant for controlling the ego-vehicle. We propose an interaction behavior prediction model based on vehicle cluster (VC) by self-attention (VC-Attention) to improve the prediction performance. Firstly, a five-vehicle based cluster structure is designed to extract the interactive features between ego-vehicle and target vehicle, such as Deceleration Rate to Avoid a Crash (DRAC) and the lane gap. In addition, the proposed model utilizes the sliding window algorithm to extract VC behavior information. Then the temporal characteristics of the three interactive features mentioned above will be caught by two layers of self-attention encoder with six heads respectively. Finally, target vehicle’s future behavior will be predicted by a sub-network consists of a fully connected layer and SoftMax module. The experimental results show that this method has achieved accuracy, precision, recall, and F1 score of more than 92% and time to event of 2.9 s on a Next Generation Simulation (NGSIM) dataset. It accurately predicts the interactive behaviors in class-imbalance prediction and adapts to various driving scenarios.

SELECTED INNOVATION TOOLS IN DESIGNING PROCESS MANAGEMENT

This contribution deals with the specification of the current immunity of strategy and status and level of procedural management in special organizational structures referred to as industrial clusters. It represents selected specifics of the design and application of innovative tools, system integration, and agile project management of clusters. It describes and explains the context between the strategy, the organizational structure, and the need for project procedural management at the current position and competitiveness of clusters on the market in demanding energy and environmental conditions. In the end, it shall submit its own author's design scheme for a new organizational project aimed at a cluster structure and draft management process and coordination of clustered projects for their clients.

Direct reaction contribution in the very low-energy 19F(p,α0) reaction

The direct reaction component of the [Formula: see text]F([Formula: see text]) reaction at [Formula: see text][Formula: see text]keV is studied for the data that became very recently available. This component is significant in this work using the direct pickup model in the framework of the DWBA formalism and indicate the strong cluster structure of [Formula: see text]F. The direct component of astrophysical S-factor is calculated for [Formula: see text]F([Formula: see text]).

Clustering and visualization of single-cell RNA-seq data using path metrics

Recent advances in single-cell technologies have enabled high-resolution characterization of tissue and cancer compositions. Although numerous tools for dimension reduction and clustering are available for single-cell data analyses, these methods often fail to simultaneously preserve local cluster structure and global data geometry. This article explores the application of power-weighted path metrics for the analysis of single cell RNA data. Extensive experiments on single cell RNA sequencing data sets confirm the usefulness of path metrics for dimension reduction and clustering. Distances between cells are measured in a data- driven way which is both density sensitive (decreasing distances across high density regions) and respects the underlying data geometry. By combining path metrics with multidimensional scaling, a low dimensional embedding of the data is obtained which respects both the global geometry of the data and preserves cluster structure. We evaluate the method both for clustering quality and geometric fidelity, and it outperforms other algorithms on a wide range of bench marking data sets.

Equivariant multiplicities of simply-laced type flag minors

Let g \mathfrak {g} be a finite simply-laced type simple Lie algebra. Baumann-Kamnitzer-Knutson recently defined an algebra morphism D ¯ \overline {D} on the coordinate ring C [ N ] \mathbb {C}[N] related to Brion’s equivariant multiplicities via the geometric Satake correspondence. This map is known to take distinguished values on the elements of the MV basis corresponding to smooth MV cycles, as well as on the elements of the dual canonical basis corresponding to Kleshchev-Ram’s strongly homogeneous modules over quiver Hecke algebras. In this paper we show that when g \mathfrak {g} is of type A n A_n or D 4 D_4 , the map D ¯ \overline {D} takes similar distinguished values on the set of all flag minors of C [ N ] \mathbb {C}[N] , raising the question of the smoothness of the corresponding MV cycles. We also exhibit certain relations between the values of D ¯ \overline {D} on flag minors belonging to the same standard seed, and we show that in any A D E ADE type these relations are preserved under cluster mutations from one standard seed to another. The proofs of these results partly rely on Kang-Kashiwara-Kim-Oh’s monoidal categorification of the cluster structure of C [ N ] \mathbb {C}[N] via representations of quiver Hecke algebras.

Effect of nanocoating morphology on the signal of X-ray Photoelectron Spectroscopy

The paper considers the application of the traditional X-ray photoelectron spectroscopy (XPS) methodology: the Overlayer Thickness Determination for the analysis of coating parameters. In particular situations considered in this work, it is energetically favorable for the atoms of the coating to form clusters, but not be evenly distributed on the surface of the substrate material. The change in the XPS signal is analyzed in situations when the coating is not a plane-parallel homogeneous layer, but an island (cluster) structure. The mathematical model of the XPS signal formation is considered for the case of the cluster covering in the form of parallelepipeds. Photoelectron path distributions (in the coating material) analysis indicated a strong dependence of the signal on the viewing angle. For the purpose of analysis, experimental spectra were obtained for several samples: gold depositions of various thicknesses on a silicon substrate. The spectra were measured for different viewing angles of photoelectrons and interpreted within the Straight Line Approximation (SLA). It is shown that proposed simplest model of an island coating allows to describe the effect of a decrease in the value of the effective average coating thickness, determined in plane-parallel geometry, with an increase in the viewing angle, observed in XPS experiments with angular resolution.

Zooplankton diversity monitoring strategy for the urban coastal region using metabarcoding analysis

Marine ecosystems in urban coastal areas are exposed to many risks due to human activity. Thus, long-term and continuous monitoring of zooplankton diversity is necessary. High-throughput DNA metabarcoding has gained recognition as an efficient and highly sensitive approach to accurately describing the species diversity of marine zooplankton assemblages. In this study, we collected 30 zooplankton samples at about 2-week intervals for 1 year. Zooplankton diversity showing a typical four season pattern. Of the “total” and “common” zooplankton, we assigned 267 and 64 taxa. The cluster structure and seasonal diversity pattern were rough when only the “common” zooplankton was used. Our study examined how to maximize the benefits of metabarcoding for monitoring zooplankton diversity in urban coastal areas. The results suggest that to take full advantage of metabarcoding when monitoring a zooplankton community, it is necessary to carefully investigate potential ecosystem threats (non-indigenous species) through sufficient curation rather than disregarding low-abundance operational taxonomic units.