Parallel Approach Research Articles

Parallel approaches for a decision tree-based explainability algorithm

While nowadays Machine Learning (ML) algorithms have achieved impressive prediction accuracy in various fields, their ability to provide an explanation for the output remains an issue. The explainability research field is precisely devoted to investigating techniques able to give an interpretation of ML algorithms’ predictions. Among the various approaches to explainability, we focus on GLEAMS: a decision tree-based solution that has proven to be rather promising under various perspectives, but suffers a sensible increase in the execution time as the problem size grows.In this work, we analyse the state-of-the-art parallel approaches to decision tree-building algorithms and we adapt them to the peculiar characteristics of GLEAMS. Relying on an increasingly popular distributed computing engine called Ray, we propose and implement different parallelization strategies for GLEAMS. An extensive evaluation highlights the benefits and limitations of each strategy and compares the performance with other existing explainability algorithms.

Creswell(1998)의 질적연구 유형 분류방식의 유용성과 한계: 교육전기 접근방법을 통한 비판적 성찰 기록

This study is a critical reflection by a qualitative researcher on the usefulness and limitations of Creswell(1998)’s five qualitative research approaches. The results of the reflection show that (1) as Creswell(1998) suggests, it is important to clearly understand the characteristics and application of different approaches, and to properly apply the appropriate qualitative research approach according to the purpose of the research. However, (2) rather than listing the five qualitative research approaches (narrative, phenomenology, ethnography, grounded theory, and case study) in parallel, as Creswell (1998) does, it is appropriate to categorize case study as a basic type of qualitative research with common characteristics that cut across the other sub-types of qualitative research. Furthermore, (3) There is a lot of overlap between the different approaches to qualitative research, so a flat description of them as simply existing in parallel is likely to confuse novice researchers. Based on the results of the discussion, the conclusion presents an alternative perspective that can compensate for the limitations of Creswell’s(1998) categorization of qualitative research approaches.

An efficient parallel approach for quad-constellation GNSS real-time precise orbit determination enabling 5-second intervals updating

Real-time Global Navigation Satellite System (GNSS) precise orbits are essential prerequisites for real-time GNSS-based applications. However, the high computational demands of filter-based real-time multi-GNSS POD methods present a significant challenge for practical utilization. This study introduces an efficient parallel approach for quad-constellation GNSS real-time POD using the square root information filter (SRIF). In the proposed approach, parallel QR factorization algorithms are implemented to expedite parameter estimation based on SRIF, while parallel processing model is designed for real-time POD to enhance its computational efficiency. Validation experiments show that the computation time for measurement and time update in SRIF can be reduced by 78.5% and 80.5%, respectively, compared to the traditional serial approach. Additionally, parallel computing improves the computational efficiency of GNSS processing algorithms in real-time POD by approximately 70%. The proposed approach enables the completion of quad-system real-time POD solutions with 90 stations within 5 s, representing an impressive 73.0% reduction.

A Comparison of Ryu and Pox Controllers: A Parallel Implementation

Software Defined Network (SDN) network controllers have limitations in handling large volumes of data generated by switches, which can slow down their performance. Using parallel programming methods such as threading, multiprocessing, and MPI aims to improve the performance of the controller in handling a large number of switches. By considering factors such as memory usage, CPU consumption, and execution time. The test results show that although RYU outperforms POX in terms of faster execution time and lower CPU utilization rate, POX shows its prowess by exhibiting less memory usage despite higher CPU utilization rate than RYU. The use of the parallel approach proves advantageous as both controllers exhibit enhanced efficiency levels. Ultimately, RYU's impressive speed and superior resource optimization capabilities may prove to be more strategic than POX over time. Taking into account the specific needs and prerequisites of a given system, this research provides insights in selecting the most suitable controller to handle large-scale switches with optimal efficiency.

New parallelism and heuristic approaches for generating tree t‐spanners in graphs

SummaryThe ‐admissibility is a min‐max problem that concerns to determine whether a graph contains a spanning tree in which the distance between any two adjacent vertices of is at most in . The stretch index of , , is the smallest for which is ‐admissible. This problem is in P for , NP‐complete for , , and remaining open for . In a very recent development, Couto et al. (Inf Process Lett, 2022; 177: 106265) introduced both sequential and parallel algorithms for constructing spanning trees. Additionally, they proposed two greedy heuristics for generating a candidate solution tree, but they left unresolved the issue of how to decide between two vertices when both have equal chances of being chosen in a greedy step. This criterion is important, since different branches can yield different stretch indexes. In response to this question, we developed nine new heuristics that use the concept of vertex importance in complex networks. Our research evaluates results on several types of graphs, including Barabási‐Albert, Erdős‐Rényi, Watts‐Strogatz, and bipartite graphs. Furthermore, we introduce a new parallel algorithm that employs a method using induced cycle of the graph to compare its performance with previously proposed algorithms. We develop a deep analysis on the proposed strategies (parallel and heuristics) comparing all of them to the other ones in the literature and as a result we obtain the best results so far in order to obtain exact values (or heuristics) of stretch indexes.

Distribution of resources beyond 5G networks with heterogeneous parallel processing and graph optimization algorithms

In this paper, a design model for resource allocation is formulated beyond 5G networks for effective data allocations in each network nodes. In all networks, data is transmitted only after allocating all resources, and an unrestrained approach is established because the examination of resources is not carried out in the usual manner. However, if data transmission needs to occur, some essential resources can be added to the network. Moreover, these resources can be shared using a parallel optimization approach, as outlined in the projected model. Further the designed model is tested and verified with four case studies by using resource allocator toolbox with parallax where the resources for power and end users are limited within the ranges of 1.4% and 6%. Furthermore, in the other two case studies, which involve coefficient determination and blockage factors, the outcomes of the proposed approach fall within the marginal error constraint of approximately 31% and 87%, respectively.

Synthesis and Characterization of BaZrS3 Thin Films via Stacked Layer Methodology: A Comparative Study of BaZrS3 on Zirconium Foil and Silicon Carbide Substrates

Chalcogenide perovskites are an emerging class of semiconductors particularly interesting for optoelectronic applications due to their high absorption coefficients and direct bandgaps of 1.7–1.9 eV. However, few processing routes have been developed for the synthesis of BaZrS3 thin films. The advances in the fabrication of BaZrS3 thin films using zirconium foil as precursor and substrate is a new strategy, which is compared with the results of BaZrS3 thin films deposited on silicon carbide (SiC) substrate. In this study, a stacked layer methodology, inspired by the fabrication methods used for chalcopyrites and kesterites, is employed. The main objective is to facilitate the conversion of these layers into the desired perovskite structure through annealing. In methodology, we initiated the process by depositing elemental Zr by sputtering on SiC, followed by BaS on top of the Zr through electron beam evaporation and then sulphurized and annealed at high temperature (≈1000–1050 °C) to form BaZrS3 in an inductively heated physical vapour transport setup in the presence of elemental sulphur. In the parallel approach the methodology remains same, with the sole deviation of zirconium foil, which act as both the substrate and the source. The successful synthesis of BaZrS3 is confirmed by X‐Ray diffraction (XRD), scanning electron microscopy (SEM), and energy‐dispersive X‐Ray spectroscopy (EDS), while the optical band gap is analyzed by UV‐Vis. The microstructure of the BaZrS3 films shows the polycrystalline structure and surface roughness.

A new hybrid parallel genetic algorithm for multi-destination path planning problem

This paper proposes a new parallel approach of multi objective genetic algorithm for path planning problem. The main contribution of this work is to reduce the population size that effect in decreasing processing times of finding the optimum path for multi destination problem. This is achieved by combining the local population of island parallel approach and global population of global parallel approach. Various experiments have been conducted to evaluate the new hybrid parallel genetic algorithm (HPGA) in solving multi-objective path planning problems. Three different test areas with 2 destinations were used to assess the performance of HPGA. Furthermore, this work compares HPGA and sequential genetic algorithm (SeqGA), as well as compared to other existing parallel genetic algorithm (GA) methods. From experimental results show that proposed HPGA outperform others, in term of processing time i.e., up to 3.6 times speedup faster, and lowest GA parameter values. This proposed HPGA can be utilized to design robots with fast and consistent path planning, especially with various obstecles.

Parallel Processing Impact on Random Forest Classifier Performance: A CIFAR-10 Dataset Study

Using the CIFAR-10 dataset, this research investigates how parallel processing affects the Random Forest method's machine learning performance. Accuracy and training time are highlighted in the study as critical performance indicators. Two cases were studied, one with and one without parallel processing. The results show the strong prediction powers of the Random Forest algorithm, which continues to analyze data in parallel while retaining a high accuracy of 97.50%. In addition, training times are notably shortened by parallelization, going from 0.6187 to 0.4753 seconds. The noted increase in time efficiency highlights the importance of parallelization in carrying out activities simultaneously, which enhances the training process's computational efficiency. These results provide important new information about how to optimize machine learning algorithms using parallel processing approaches.

Low-frequency Radio Interferometric Array Imaging Pipeline Optimization Based on Distributed Execution Framework

The Square Kilometre Array (SKA) project is an international collaboration to build the world’s largest radio telescope, whose sensitivity and measurement speed will be an order of magnitude higher than those of all current radio telescopes. Radio continuum survey is one of the main observation mode of the SKA, and the establishment of a standard map of the survey area based on continuum imaging will provide an important foundation for subsequent astronomical science. The GaLactic and Extragalactic All-sky Murchison Widefield Array survey eXtended (GLEAM-X) is a project of the SKA pilot telescope Murchison Widefield Array (MWA) in 2018—2020. GLEAM-X is a new radio continuum survey project to be carried out with the MWA Phase II expansion array in 2018—2020. The experience of optimizing the imaging pipeline based on the distributed execution framework will help to solve the problem of massive data processing. In this paper, we describe the process steps of GLEAM-X imaging pipeline, integrate and improve it, and realize parallel processing of multiple pipelines on the China SKA Regional Centre Prototype (CSRC-P), and verify the deployment and test the correctness of the imaging pipeline system using GLEAM-X observation data. The GLEAM-X observations were used to validate the deployment of the imaging pipeline system and test its correctness. Then, to optimize the pipelines and improve the processing efficiency, the Data Activated Liu Graph Engine (DALiuGE) was used to integrate the imaging pipelines into the DALiuGE execution framework to automate the distributed parallel processing of the pipelines. Performance tests and results’ analysis show that the optimized imaging pipeline based on the DALiuGE execution framework has better performance, more flexible adaptability, and scalability than the traditional parallel approach, and can support future large-scale continuum imaging experiments during the first phase of SKA commissioning.

A scalable multi-density clustering approach to detect city hotspots in a smart city

In the field of Smart City applications, the analysis of urban data to detect city hotspots, i.e., regions where urban events (such as pollution peaks, virus infections, traffic spikes, and crimes) occur at a higher density than in the rest of the dataset, is becoming a common task. The detection of such hotspots can serve as a valuable organizational technique for framing detailed information about a metropolitan area, providing high-level spatial knowledge for planners, scientists, and policymakers. From the algorithmic viewpoint, classic density-based clustering algorithms are very effective in discovering hotspots characterized by homogeneous density; however, their application on multi-density data can produce inaccurate results. For such a reason, since metropolitan cities are characterized by areas with significantly variable densities, multi-density clustering approaches are more effective in discovering city hotspots. Moreover, the growing volumes of data collected in urban environments require the development of parallel approaches, in order to take advantage of scalable executions offered by Edge and Cloud environments. This paper describes the design and implementation of a parallel multi-density clustering algorithm aimed at analyzing high volumes of urban data in an efficient way. The experimental evaluation shows that the proposed parallel clustering approach takes out encouraging advantages in terms of execution time, speedup, and efficiency.

Optimizing Enhanced Extended Topological Active Nets Model Using Parallel Processing

In numerous clinical applications that support the diagnosis and treatment planning of a broad variety of disorders, medical image segmentation is essential. Medical picture segmentation using the Enhanced Extended Topological Active Net (EETAN) model has proven to be successful in correctly identifying structures. This study suggests a novel way to combine the best clustering techniques and parallel processing approaches to maximize the segmentation performance of the EETAN model. The Probabilistic Depth Search Optimization (PDSO) Algorithm, which makes the parallel searching technique to find the ideal contour set, is responsible for this. This work implements parallel processing and ideal clustering to improve the EETAN model's performance in medical image segmentation. Performance metrics like accuracy, precision, recall, dice similarity, and computational time are used for a comparison study. The results demonstrate the notable enhancements attained by employing parallel processing and effective clustering.

A novel proteomic signature of osteoclast differentiation unveils the deubiquitinase UCHL1 as a necessary osteoclastogenic driver

Bone destruction, a major source of morbidity, is mediated by heightened differentiation and activity of osteoclasts (OC), highly specialized multinucleated myeloid cells endowed with unique bone-resorptive capacity. The molecular mechanisms regulating OC differentiation in the bone marrow are still partly elusive. Here, we aimed to identify new regulatory circuits and actionable targets by comprehensive proteomic characterization of OCgenesis from mouse bone marrow monocytes, adopting two parallel unbiased comparative proteomic approaches. This work disclosed an unanticipated protein signature of OCgenesis, with most gene products currently unannotated in bone-related functions, revealing broad structural and functional cellular reorganization and divergence from macrophagic immune activity. Moreover, we identified the deubiquitinase UCHL1 as the most upregulated cytosolic protein in differentiating OCs. Functional studies proved it essential, as UCHL1 genetic and pharmacologic inhibition potently suppressed OCgenesis. Furthermore, proteomics and mechanistic dissection showed that UCHL1 supports OC differentiation by restricting the anti-OCgenic activity of NRF2, the transcriptional activator of the canonical antioxidant response, through redox-independent stabilization of the NRF2 inhibitor, KEAP1. Besides offering a valuable experimental framework to dissect OC differentiation, our study discloses the essential role of UCHL1, exerted through KEAP1-dependent containment of NRF2 anti-OCgenic activity, yielding a novel potential actionable pathway against bone loss.

Navigating parallel interactive narratives in virtual reality

ABSTRACT The application of Campbell’s hero’s journey in cinema commonly adopts a director-centric approach in which the director decides how the viewer will see the narrative. A parallel interactive narrative in virtual reality titled Table for Two adopts screen production enquiry to investigate the divergence from the hero’s journey from a director-centric approach towards a navigator-centric approach. Here, the conventional cinema viewer is proposed as a navigator due to the immersive, interactive, and intellectual role they perform in terms of 360-degree viewing, teleportation, and meaning creation. While hero’s journey has been examined across various narrative platforms such as literature, theatre, cinema, gaming, theme parks, education, and more; there are limited discussions on the hero’s journey and its relevance to virtual reality experiences. Table for Two is a one-shot 360-degree audio-visual experience captured simultaneously by three 360-degree cameras. The navigator is encouraged to teleport from one narrative to another as the remaining narratives unfold irrespective of their choices. Each navigator is empowered to untie the knotted pieces and construct a narrative puzzle. Table for Two diverges from the hero’s journey through its multi-linear structure, simultaneous parallel narrative approach, and 360-degree spatial storytelling that offers freedom to navigate.

Accelerating multivariate functional approximation computation with domain decomposition techniques

Modeling large datasets through Multivariate Functional Approximations (MFA) provide an elegant way to handle many visualization and scientific analysis workflows. The process necessitates scalable data partitioning methods to compute MFA representations efficiently without compromising the accuracy or continuity of the reconstructed solution. We propose a domain-decomposed method for computing the MFA with B-spline bases, which reduces the total work per task and uses a restricted Additive Schwarz (RAS) method to converge the control point data degrees-of-freedom along subdomain boundaries. We provide an in-depth analysis of the parallel approach with domain decomposition solvers, aiming to minimize local subdomain error residuals and recover high-order continuity at subdomain interfaces with appropriate choices of knot overlaps. The communication cost, determined by the overlap regions in the RAS implementation, is optimized to recover the numerical error profile of the single subdomain case. Our proposed method stands in contrast to previous methods, which typically only recover either C0 or at best C1 continuity for arbitrary B-spline degree expansions, or those that require post-processing to blend discontinuities in the reconstructed data. We demonstrate the effectiveness of our approach using analytical and real-world datasets in 1D, 2D, and 3D through both strong and weak scaling studies. The performance results indicate that the overall cost of computing the approximation is directly proportional to the underlying nearest-neighbor communication implementation, and is only weakly dependent on the overlap region size that determines the size of the messages. This finding underscores the efficiency and scalability of our proposed method, making it a promising solution for handling large datasets in scientific workflows.

Abstract 2928: Solving the cancer mutation conundrum: A single cell, massively parallel approach for cancer mutation discovery, genome modelling and functional characterization

Abstract Large scale genomic studies are cataloguing tens of thousands of cancer mutations across a wide spectrum of tumor types. Conventional DNA sequencing methods have limitations in terms of identify mutations at high resolution among single cells. Moreover, determining cancer mutations’ phenotype and function remains an enormous challenge - most mutations are assigned to the category of variants of unknown significance. In-silico functional predictions of cancer mutations are frequently used as a stopgap solution. However, these computational methods lack the ability to make refined biological and functional predictions. We demonstrated a new single cell experimental strategy that accelerates the discovery of cancer mutations and their characterization. This genomic technology framework does the following: (1) identify disease-related mutations at the resolution of individual cells from tumors or cell lines; (2) genome engineer these candidate mutations into cellular experimental systems for further biological characterization. First, we developed a single cell approach to identify cancer somatic mutations and even gene fusion rearrangements found in coding regions of mRNAs. Primary tumors are processed into single cell RNA-seq libraries and analyzed with nanopore long read sequencing. Across thousands of cells from primary tumors, we identified single cell mutations from the nanopore long reads that covered cancer gene transcripts. This approach allows us to even identify cancer rearrangements at single cell resolution. Next, we used short-read single cell transcriptomes to characterize the cell type and determine the gene expression characteristics of those cells with mutations. For single cell phenotypic studies, we developed a highly multiplexed technology using CRISPR base editors to directly engineer hundreds of mutations into their original genes among single cells and determine their transcriptional phenotype. Among specific individual cells, nanopore-based long-read sequencing identifies these modelled mutations directly from a target’s transcript sequence. Then, we integrate this single cell mutation data with the short-read transcriptome profile from the same individual cell. Therefore, we determine both the genotype and phenotype of a given engineered mutation from an individual cell. In summary, our work demonstrates a new, direct and highly scalable method for identifying, modelling and functionally assessing cancer mutation phenotype at single cell resolution. This approach enables the broader assessment of cancer mutations impact on tumor biology and clinical ramifications. Citation Format: Susan M. Grimes, Heonseok Kim, Sharmili Roy, Anuja Sathe, Billy T. Lau, Hanlee P. Ji. Solving the cancer mutation conundrum: A single cell, massively parallel approach for cancer mutation discovery, genome modelling and functional characterization [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2928.

Adaptive quadtree splitting parallelization (AQSP) algorithm for the VVC standard

The Versatile Video Coding (VVC) standard, also known as H.266, was released in 2020 as the natural successor to the High Efficiency Video Coding (HEVC) standard. Among its innovative coding tools, VVC extended the concept of quadtree (QT) splitting to the multi-type tree (MTT) structure, introducing binary and ternary partitions to enhance HEVC’s coding efficiency. While this brought significant compression improvements, it also resulted in a substantial increase in encoding time, primarily due to VVC’s larger Coding Tree Unit (CTU) size of 128×\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ imes $$\\end{document}128 pixels. To mitigate this, this work introduces a flexible parallel approach for the QT traversal and splitting scheme of the VVC encoder, called adaptive quadtree splitting parallelization (AQSP) algorithm. This approach is based on the distribution of coding units (CUs) among different threads using the current depth level of the QT as a basis to minimize the number of broken dependencies. In this way, the algorithm achieves a good trade-off between time savings and coding efficiency. Experimental results show that, when compared with the original VVC encoder, AQSP achieves an acceleration factor of 2.04×\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ imes $$\\end{document} with 4 threads at the expense of a low impact in terms of BD rate. These outcomes position AQSP competitively in comparison with other state-of-the-art approaches.

A parallel approach for ultra-fast state estimation in large power system using graph partitioning theory

This paper introduces a novel approach for multi-area state estimation in large transmission networks through the application of graph partitioning theory. By harnessing the eigenvalues and eigenvectors of the Laplacian matrix, a large-scale transmission network is partitioned into manageable sections. Within these partitions, state estimation processes run in parallel, markedly improving efficiency compared to conventional methods. Linear state estimation is employed within each area, expediting computations and making it adaptable to large-scale networks, which traditionally pose computational challenges. The method's efficacy is demonstrated through comprehensive validation, commencing with small networks and extending to real-world applications on the IEEE 118-bus test system and the 9241-bus European high-voltage transmission network. In comparison to the integrated network method, our approach has achieved state estimation answers with reduced computation time. The partitioning of the integrated network into multi areas has effectively mitigated computational loads, showcasing its potential for enhancing operational efficiency and reliability in complex power transmission systems. This approach not only offers a robust solution for state estimation but also represents a significant stride toward advancing the field of state estimation, promising to bolster the stability and performance of modern power grids.

Multi-objective multiverse optimization for optimal allocation of distributed energy resources: The optimal parallel processing schemes

Utilizing Distributed Generators (DGs) and Energy Storage Systems (ESSs) enhances power system reliability, drawing significant research attention. However, these systems pose challenges, compelling scientists to explore optimization methods. Our paper presents an innovative solution, Parallel Multi-Objective Multi-Verse Optimization (PMOMVO), aimed at optimizing DGs and Battery Energy Storage Systems (BESSs) allocation. This optimization addresses voltage violations and operation costs, crucial concerns for system operators and consumers. By leveraging a parallel approach, PMOMVO significantly accelerates the optimization process. We compared its results with a base case scenario, demonstrating the superior efficiency of our parallel method. It not only enhances the optimization performance but also proves its efficacy by generating optimal solutions from the Pareto front set. This research showcases the benefits of PMOMVO, offering a faster, more efficient, and reliable way to optimize power systems, benefiting both operators and consumers.

Investigating the Efficacy and Sustainability of Virtual Education Courses

As teacher educators prepare for the future of teacher education, it is important to address the sustainability of virtual education courses. This study used a pre-existing model provided by the New York State Education Department as a framework to define the types of courses that may appear in an education program. Data was collected using a convergent parallel mixed methods approach to simultaneously gather qualitative and quantitative data. A total of 640 education courses from accredited education programs across the United States were analyzed. Findings indicated the type of course did not play a role in the course effectiveness during the pandemic, suggesting all education courses can be delivered effectively in a virtual modality. The delivery method of undergraduate courses was indicated by participants to be less effective during the pandemic than graduate courses. Results suggest courses that involve fieldwork (i.e., Field Experience, Student Teaching, or Practicum) were indicated by participants to be the least sustainable as virtual courses post-pandemic. Results also suggest that although virtual education courses at the graduate level are sustainable in education programs post-pandemic, undergraduate courses may not be. Advantages to virtual course delivery clustered around logistics and access while challenges clustered around engagement and hands-on-learning. This study provides curriculum guidance to those across the globe in the position of making modality decisions.