Graph Processing Research Articles

Biological gene extraction path based on knowledge graph and natural language processing.

The continuous progress of society and the vigorous development of science and technology have brought people the dawn of maintaining health and preventing and controlling diseases. At the same time, with the update and iteration of bioinformatics technology, the current biological gene research has also undergone revolutionary changes. However, a long-standing problem in genetic research has always plagued researchers, that is, how to find the most needed sample genes from a large number of sample genes, so as to reduce unnecessary research and reduce research costs. By studying the extraction path of biological genes, it can help researchers to extract the most valuable research genes and avoid wasting time and energy. In order to solve the above problems, this paper used the Bhattacharyya distance index and the Gini index to screen the sample genes when extracting the characteristic genes of breast cancer. In the selected 49 public genes, 6 principal components were extracted by principal component analysis (PCA), and finally the experimental results were tested. It was found that when the optimal number of characteristic genes was selected as 5, the recognition rate of genes reached the highest 90.31%, which met the experimental requirements. In addition, the experiment also proved that the characteristic gene extraction method designed in this paper had a removal rate of 99.75% of redundant genes, which can greatly reduce the time and money cost of research.

An unsupervised learning-guided multi-node failure-recovery model for distributed graph processing systems

An unsupervised learning-guided multi-node failure-recovery model for distributed graph processing systems

Estimation of high-dimensional vector autoregression via sparse precision matrix

Summary We consider the problem of estimating sparse vector autoregression (VAR) via penalized precision matrices. This matrix is the output of the underlying directed acyclic graph of the VAR process, whose zero components correspond to the zero coefficients of the graphical representation of the VAR. The sparsity-based precision matrix estimator is deduced from the D-trace loss with convex and nonconvex penalty functions. We establish the consistency of the penalized estimator and provide the conditions for which all true zero entries of the precision matrix are actually estimated as zero with probability tending to one. The relevance of the method is supported by simulated experiments and a real data application.

Bet-GAT: An Efficient Centrality-Based Graph Attention Model for Semi-Supervised Node Classification

Graph Neural Networks (GNNs) have witnessed great advancement in the field of neural networks for processing graph datasets. Graph Convolutional Networks (GCNs) have outperformed current models/algorithms in accomplishing tasks such as semi-supervised node classification, link prediction, and graph classification. GCNs perform well even with a very small training dataset. The GCN framework has evolved to Graph Attention Model (GAT), GraphSAGE, and other hybrid frameworks. In this paper, we effectively usd the network centrality approach to select nodes from the training set (instead of a traditional random selection), which is fed into GCN (and GAT) to perform semi-supervised node classification tasks. This allows us to take advantage of the best positional nodes in the network. Based on empirical analysis, we choose the betweenness centrality measure for selecting the training nodes. We also mathematically justify why our proposed technique offers better training. This novel training technique is used to analyze the performance of GCN and GAT models on five benchmark networks—Cora, Citeseer, PubMed, Wiki-CS, and Amazon Computers. In GAT implementations, we obtain improved classification accuracy compared to the other state-of-the-art GCN-based methods. Moreover, to the best of our knowledge, the results obtained for Citeseer, Wiki- CS, and Amazon Computer datasets are the best compared to all the existing node classification methods.

Graphfire: Synergizing Fetch, Insertion, and Replacement Policies for Graph Analytics

Despite their ubiquity in many important big-data applications, graph analytic kernels continue to challenge modern memory hierarchies due to their frequent, long-latency, pointer indirect accesses to vertex property data. Such accesses exhibit poor locality and variable reuse that trouble cache replacement policies, and consequently increase memory bandwidth pressure. Specialized graph-tailored prefetching mechanisms, processor designs, and memory hierarchy engines have been developed to tolerate the long latencies of such accesses. However, these approaches are either too bandwidth-intensive, require invasive hardware changes that inhibit general-purpose computation flexibility, or rely on software preprocessing that limits true speedup. This work introduces Graphfire, a flexible memory hierarchy approach that learns different access patterns in graph processing and exploits the synergy of specialized fetch, insertion, and replacement optimizations for problematic indirect accesses without relying on software or ISA support. More specifically, Graphfire identifies when these irregular accesses occur and employs tailored access granularities, data-aware insertion, and frequency-based replacement accordingly. It achieves up to a 1.79× speedup (geomean 1.3×) and these improvements scale due to bandwidth efficiency; with 64 cores, Graphfire yields up to a 71.33× speedup (geomean 63.32×) over a single baseline core and allows memory-bound graph analytic codes to scale far beyond prior work.

Multi-Level Indoor Path Planning and Clearance-Based Path Optimization for Search and Rescue Operations

In this study a multi-level path planning system is proposed for indoor search and rescue operations. Requirements for the path planning system are derived based on the operational concept of the integrated indoor navigation system. Different aspects of various path planning algorithms are assessed, and their suitability to search and rescue operations in structured indoor environments is investigated. A five-step path planning system is proposed, which consists of map pre-processing, segment path planning, graph processing, route optimization, and path post-processing. The proposed method addresses a multi-goal path planning problem in a multi-story building in a computationally efficient way by adopting a graph-based approach while satisfying such requirements as clearance conditions in the pre- and post-processing steps. Furthermore, a multi-query approach is adopted to exploit the response time and earn flexibility with respect to environmental changes. The effectiveness of the proposed path planning system is demonstrated through numerical simulations.

Исследование различных подходов к управлению двухмассовыми объектами

The reasons for elastic oscillations occurring in the mechanisms of multi-mass objects are considered. Various control systems for complex two-mass objects are analyzed. A mathematical description of a twomass system interconnected through an elastic drive belt is provided. According to the given system of equations, a block diagram of the object under consideration has been constructed in the mathematical environment of MatLab and Simulink. The synthesis and comparative analysis is conducted for the following control systems for a two-mass object: slave control with external feedback on the speed of the first mass; multiloop slave regulation; modal control; subordinate regulation using a correction factor and a neurocontroller with a reference model. The resulting graphs of transient processes in terms of moment and speed of the driven pulley of the considered twomass system with an elastic replacement belt are presented and analyzed. The simulation results have shown that control systems with a modal controller and a neurocontroller show good results in the problem of damping elastic vibrations of a two-mass system in comparison with other considered control systems.

Теоретические исследования гидросистемы рулевого управления строительных и дорожных машин с усилителями потока

The article presents the results of theoretical studies of the functioning of flow amplifiers, the purpose of which is to increase the useful power of the hydraulic steering systems of heavy construction and road machines, which is especially important when upgrading existing equipment and designing a new one. Flow amplifiers allow to increase the supply of working fluid to the hydraulic cylinders for turning a multiple of the dosed volume by the dosing pump. A distinctive feature of the study is the consideration of two flow amplifiers of throttling and volumetric types, working in pairs. The design, the principle of operation are described, the main calculated dependencies are presented, on the basis of which mathematical modeling was carried out. To evaluate the qualitative and quantitative characteristics of the amplifiers, graphs of transient processes in the steering hydraulic system, as well as the static characteristics of the flow amplifiers, were obtained. The analysis of the obtained dependencies showed acceptable accuracy and quality of the hydraulic system. The results obtained can be used in the synthesis of such systems.

РАЗРАБОТКА СИСТЕМЫ УПРАВЛЕНИЯ ПОГРУЖНОГО ВЕНТИЛЬНОГО ЭЛЕКТРОПРИВОДА ЭЛЕКТРОБУРА

Due to the active development of the oil and gas service, the construction of oil and gas wells is becoming increasingly important for the economy of Russia. This work is aimed at developing a mathematical description and model of the control system of the valve electric drive of an electric drill, as well as analyzing the transient processes of the obtained system. The work presents a mathematical description of the dynamic modes of operation of the electric drive system with a synchronous machine with permanent magnets, and a structural diagram is built. Using the Matlab Simulink environment, a model of the system was built. Graphs of transient processes during start-up and load changes were obtained, based on which conclusions were drawn about the operation of the electric drill control system. It was found that the control system is capable of adapting and maintaining the speed at a given level when the load on the shaft and the control signal change.

Моделирование полета беспилотного летательного аппарата в Matlab Simulink

In this article, a simulation model of an unmanned aerial vehicle in Matlab Simulink has been developed. To build a simulation model, a mathematical description of the equations of motion of an aircraft (AC) is performed: kinematic and dynamic models with six degrees of freedom and twelve state variables, and a mathematical description of external forces acting on the aircraft is also performed. The simulation model includes several blocks that interact with each other to ensure the operation of the unmanned aerial vehicle: the air craft dynamics unit, which is responsible for simulating the movement of the aircraft in three dimensions, the autopilot unit, which simulates the operation of the aircraft navigation system and provides stabilization in the air, the control unit engines, which regulates the speed of rotation of the engines depending on the control signals, an external environment block that simulates the air environment in which the aircraft moves. The main advantage of the constructed model is the ability to change the parameters of the aircraft interactively and evaluate the main parameters using graphs. The correctness of the autopilot operation is confirmed by the graphs of transient processes along the three axes of the aircraft: roll, pitch, yaw. In the future, based on the constructed model, it is possible to create a fully automatic obstacle avoidance system that will allow the unmanned aerial vehicle to safely move from the start to the end point, taking into account all the obstacles in the flight path. Thus, the simulation model of an unmanned aerial vehicle based on Simulink makes it possible to evaluate the operation of the autopilot and other aircraft systems with high accuracy and detail. Its use can be useful in the design and testing of new unmanned aerial vehicles.

Superpixelwise Low-Rank Approximation-Based Partial Label Learning for Hyperspectral Image Classification

Insufficient prior knowledge of a captured hyperspectral image (HSI) scene may lead the experts or the automatic labeling systems to offer incorrect labels or ambiguous labels (i.e., assigning each training sample to a group of candidate labels, among which only one of them is valid; this is also known as partial label learning) during the labeling process. Accordingly, how to learn from such data with ambiguous labels is a problem of great practical importance. In this paper, we propose a novel superpixelwise low-rank approximation (LRA)-based partial label learning method, namely SLAP, which is the first to take into account partial label learning in HSI classification. SLAP is mainly composed of two phases: disambiguating the training labels and acquiring the predictive model. Specifically, in the first phase, we propose a superpixelwise LRA-based model, preparing the affinity graph for the subsequent label propagation process while extracting the discriminative representation to enhance the following classification task of the second phase. Then to disambiguate the training labels, label propagation propagates the labeling information via the affinity graph of training pixels. In the second phase, we take advantage of the resulting disambiguated training labels and the discriminative representations to enhance the classification performance. The extensive experiments validate the advantage of the proposed SLAP method over state-of-the-art methods.

Multiscale Deep Graph Convolutional Networks for Intelligent Fault Diagnosis of Rotor-Bearing System Under Fluctuating Working Conditions

The rotor-bearing system is widely used in various high-end electro-hydraulic equipment, which provides specific support, rotation, and other integral functions. However, the fluctuating working conditions of the rotor-bearing system will cause more significant disordered fluctuations in the measured signals. This article proposes a new algorithm called multiscale deep graph convolutional networks (MS-DGCNs) to alleviate this problem. The designed MS-DGCNs algorithm combines a new multiscale intra-class fine coarse-grained processing and multiscale graph convolution kernels. Accordingly, an intelligent fault diagnosis method based on MS-DGCNs for the rotor-bearing system under fluctuating conditions is designed to learn more feature representations and accuracy. First, a sliding window is employed to divide the collected vibration signals into a series of subsignals. The multiscale signal processing is performed to obtain different degrees of the fine-coarse time series. Then, a graph convolution with the multiscale convolution kernel is designed. Finally, the soft-max classifier is combined for intelligent fault diagnosis. The experimental results of the double-span rotor-bearing system under fluctuating conditions well demonstrate that the method has the higher accuracy and generalization.

Multi-Mode SpMV Accelerator for Transprecision PageRank With Real-World Graphs

With the development of Internet networks, the PageRank algorithm, which was initially developed to recommend important pages in Google’s web search systems, is widely used as the basis of various ranking systems in graph processing fields. However, PageRank algorithm requires Sparse Matrix-Vector Multiplication (SpMV) repeatedly which becomes main bottleneck for the calculation. In this study, we present the multi-mode SpMV accelerator for half-to-single transprecision PageRank with real-world graphs. To support transprecision, where the operation performs in half-precision (FP16) initially and changes its precision to single-precision (FP32), the proposed multi-mode SpMV accelerator can perform both dual FP16 mode and single FP32 mode. In dual FP16 mode, the proposed accelerator performs two FP16 SpMV in parallel, and in single FP32 mode, it performs one FP32 SpMV with the same hardware resources. Also, for the reduction of memory footprint, the proposed accelerator supports the Compressed Sparse Row (CSR) format. In addition, dual-issue accumulator and multi-mode transprecision multiplier are presented to support both FP16 and FP32 modes. Validation of the proposed transprecision PageRank algorithm is performed with four real-world graph datasets, resulting in a low 0-4% error rate and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$1.3\times $ </tex-math></inline-formula> - <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$1.9\times $ </tex-math></inline-formula> speedup compared to single-precision PageRank computation without transprecision.

MapReduce for Graphs Processing: New Big Data Algorithm for 2-Edge Connected Components and Future Ideas

MapReduce for Graphs Processing: New Big Data Algorithm for 2-Edge Connected Components and Future Ideas

Application of Graphs in a One Health Framework.

The One Health framework, which advocates the crucial interconnection between environmental, animal, and human health and well-being, is becoming of increasing importance and acceptance in health sciences over the last years. The hottest public health topics of the latest years, like zoonotic diseases (e.g., the recent pandemic) or the increasing antibiotic resistance, characterized by many as "pandemic of the future," make the more holistic and combinatorial approach of One Health a necessity to combat such complex problems. Multiple graphs and graph theory have found applications in health sciences for many years, and they can now extend to usage across all levels of a One Health approach to health, ranging from genome, one disease level, to epidemiology and ecosystem graphs. For that last ecosystem layer, a proposed approach is the utilization of process graphs from the chemical engineering field, in order to understand a whole system and what constitute the most crucial aspects of a One Health issue in ecosystem level. Here P-graphs are focused alongside their combinatorial algorithms, implemented in R, and their application researched in an effort to extract information and plan interventions.

Methods for improving the efficiency of a wind power plant and reliable estimation of output control parameters and disturbances

A method for controlling a wind power plant based on the use of the optimization model that takes account of the drive vibration loading under different operating modes of the power unit is validated. The transfer functions of the rotor systems dynamic links are proposed, taking account of the influence of wind speed and electrical load, the block diagram of the control system and the transient process graph of the angular velocity regulating of the installation generator rotor. To ensure the reliability of the output control parameters and disturbing influences estimation, the determining method substantiation of the dynamic links correlation coefficients for the wind power plant rotor system is performed. These coefficients are necessary to develop models of the system dynamic behavior used in the modification of the wind power plant automated control in order to reduce the vibrations of all rotor systems elements under different operating modes of the power unit that will improve the reliability of modern wind turbines. The correlation function parameters to determine the correlation dependences for the measured values of wind speed and electrical load are found. Formulas are derived for determining the spectral density of input and output random processes at changing wind speed and electrical load, as well as the mutual spectral density of input and output random processes for dynamic links of rotor systems (rotary blades, rotor, generator, gearbox) based on the block diagram of the entire process in order to determine the correlation degree of transfer functions. The analysis of the input and output signals of the transfer functions with respect to their coherence is carried out from a correlation coefficients calculation of the rotor systems dynamic links. This makes it possible to describe the transfer function within the framework of the vibration fields analysis for the unit as a whole. Keywords wind power plant, control, rotor system, vibration, correlation coefficient

Simulation and analysis of transition processes in linear electrical circuits using MatLab software package

The research work considers the simulation and analysis of transition processes in linear electrical circuits using the MatLab software package. The transition processes occurring in the following linear electrical circuits were analyzed: RL equivalent circuit in case of connecting the relay control winding to a DC voltage source; RC equivalent circuit in case of connecting a capacitor through a limiting resistor to a DC voltage source and disconnecting it from the source; RLC equivalent circuit in case of capacitor discharge to an inductor; this capacitor shall be pre-charged from a DC voltage source; a branched electrical circuit of the first order where its topology changes after switching. As a result of the research, the transition process graphs were obtained for which the simulated completion time of the transient process does not differ from the calculated values with an error of less than 2%. The analysis of transition processes revealed the following: the winding inductance effect on the time of relay switching; the effect of the additional resistance value on the charge (discharge) of the capacitor; the effect of the resistance value change when the capacitor is discharged in a circuit with two reactive elements; the effect of changes in the topology of the electrical circuit on the intensity of the transient process after switching.

Edge detection and graph neural networks to classify mammograms: A case study with a dataset from Vietnamese patients

Mammograms are breast X-ray images and they are used by doctors, among other purposes, as an effective means of detecting breast cancer. Screening mammography is crucial since it allows doctors to understand better the situation and have suitable intervention. The classification of medical modalities is a prerequisite for development of computer-aided diagnosis tools in healthcare, and various techniques have been proposed to automatically classify from mammography images. Though there have been several tools developed, they have been mostly validated with data collected from Western women. Based on our initial investigations, breast anatomy in Vietnamese women differs from that of Western women, due to denser breast tissue. In this paper, we propose MammoGNN – a practical solution to the classification of mammograms using the synergy between image processing techniques and graph neural networks. First, a well-founded edge detection algorithm was applied to provide input for the recommendation engine. Afterward, we empirically experimented to select suitable graph neural networks to manage the training and prediction. A mammogram dataset was curated from 2,351 Vietnamese women to validate the conceived tool. By several testing instances, MammoGNN obtains a maximum accuracy of 100%, precision and recall of 1.0 on independent and shuffle test sets for both classification of BI-RADS scores and breast density types. The experimental results also demonstrate that our proposed approach obtains an optimal prediction performance on the considered datasets, outperforming different baselines. We anticipate that the proposed approach can be deployed as a non-invasive pre-screening tool to assist doctors in performing their diagnosis activities.

Near-memory Computing on FPGAs with 3D-stacked Memories: Applications, Architectures, and Optimizations

The near-memory computing (NMC) paradigm has transpired as a promising method for overcoming the memory wall challenges of future computing architectures. Modern systems integrating 3D-stacked DRAM memory can be leveraged to prevent unnecessary data movement between the main memory and the CPU. FPGA vendors have started introducing 3D memories to their products in an effort to remain competitive on bandwidth requirements of modern memory-intensive applications. Recent NMC proposals target various types of data processing workloads such as graph processing, MapReduce, sorting, machine learning, and database analytics. In this article, we conduct a literature survey on previous proposals of NMC systems on FPGAs integrated with 3D memories. By leveraging the high bandwidth offered from such memories together with specifically designed hardware, FPGA architectures have become a competitor to GPU solutions in terms of speed and energy efficiency. Various FPGA-based NMC designs have been proposed with software and hardware optimization methods to achieve high performance and energy efficiency. Our review investigates various aspects of NMC designs such as platforms, architectures, workloads, and tools. We identify the key challenges and open issues with future research directions.

An Internet of Things based approach for Smart Home Management

The advent of the Internet of Things (IoT) paradigm has led to the construction of technological environments based on device communication. In this scenario, solutions have been developed for the intelligent management of various services, such as those related to Smart Cities. In such a scenario, widespread use is that of Smart Homes, environments that can provide management help and intelligent and tailored services to users. This paper aims to present an Internet of Things-based approach to Smart Home management. The proposed approach, which leverages graph approaches, enables the extraction and processing of contextual data, collected from sensors, and probabilistic approaches able to provide valuable information and autonomous actions. The system training phase is performed using graph approaches, which could represent an increase in Context and Situation Awareness. In the experimental phase, a prototype is used to test the proposed method’s energysaving and ventilation management capabilities, which shows encouraging findings.