Search

The article discusses the Earth Cloud, Aerosol and Radiation Explorer (EarthCARE) satellite mission, which is a joint effort between the European Space Agency (ESA) and the Japan Aerospace Exploration Agency (JAXA). The mission aims to improve weather forecasts and climate projections by studying cloud, aerosol, and radiation processes. EarthCARE will use novel instruments to retrieve more accurate properties of clouds, aerosols, and precipitation, and provide in-orbit verification of their impact on radiative fluxes. The article also pays tribute to Dr. Tobias Wehr, the mission scientist who passed away in 2023 and played a crucial role in the development of EarthCARE. [Extracted from the article]

The race for the most efficient, accurate, and universal algorithm in scientific computing drives innovation. At the same time, this healthy competition is only beneficial if the research output is actually comparable to prior results. Fairly comparing algorithms can be a complex endeavor, as the implementation, configuration, compute environment, and test problems need to be well‐defined. Due to the increase in computer‐based experiments, new infrastructure for facilitating the exchange and comparison of new algorithms is also needed. To this end, we propose a benchmark framework as a set of generic specifications for comparing implementations of algorithms using test cases native to a community. Its value lies in its ability to fairly compare and validate existing methods for new applications, as well as compare newly developed methods with existing ones. As a prototype for a more general framework, we have begun building a benchmark tool for the model order reduction (MOR) community. The data basis of the tool is the collection of the Model Order Reduction Wiki (MORWiki). The wiki features three main categories: benchmarks, methods, and software. An editorial board curates submissions and patrols edited entries. Data sets for linear and parametric‐linear models are already well represented in the existing collection. Data sets for non‐linear or procedural models, for which only evaluation data, or codes/algorithmic descriptions, rather than equations, are available, are being added and extended. Properties and interesting characteristics used for benchmark selection and later assessments are recorded in the model metadata. Our tool, the Model Order Reduction Benchmarker (MORB), is under active development for linear time‐invariant systems and solvers. An ontology (MORBO) and knowledge graph are being developed in parallel. They catalog benchmark problem sets and their metadata and will also be integrated into the Mathematical Research Data Initiative (MaRDI) Portal to help improve the findability of such data sets. MORB faces a number of technical and field‐specific challenges, and we hope to recruit community input and feedback while presenting some initial results. [ABSTRACT FROM AUTHOR]

The augmented Lagrangian method (ALM) is classic for canonical convex programming problems with linear constraints, and it finds many applications in various scientific computing areas. A major advantage of the ALM is that the step for updating the dual variable can be further relaxed with a step size in (0,2), and this advantage can easily lead to numerical acceleration for the ALM. When a separable convex programming problem is discussed and a corresponding splitting version of the classic ALM is considered, convergence may not be guaranteed and thus it is seemingly impossible that a step size in (0,2) can be carried on to the relaxation step for updating the dual variable. We show that for a parallel splitting version of the ALM, a step size in (0,2) can be maintained for further relaxing both the primal and dual variables if the relaxation step is simply corrected by a rank-two matrix. Hence, a rank-two relaxed parallel splitting version of the ALM with a step size in (0,2) is proposed for separable convex programming problems. We validate that the new algorithm can numerically outperform existing algorithms of the same kind significantly by testing some applications. [ABSTRACT FROM AUTHOR]

It is particularly important to plan and design the rural landscape of the Qiang nationality, but the regional characteristics of the rural landscape of the Qiang nationality in the upper reaches of the Minjiang River must be maintained at the same time. On the basis of the regional culture formed by the regional geography and cultural environment of the Qiang ethnic group in the upper reaches of the Minjiang River, combined with the investigation of the Qiang ethnic group in the upper reaches of the Minjiang River, this study explores the formal characteristics and formation reasons of the rural landscape of the Qiang ethnic group in the upper reaches of the Minjiang River from the perspective of rural landscape planning and design. It summarizes the formal features of the Qiang rural landscape in the upper reaches of the Minjiang River in a graphical language. It analyzes the problems faced by traditional rural landscapes. It then summarizes the experience and deficiencies in practice through the research on the existing Qiang rural landscape planning and design. It analyzes the influence of the regional culture of the Qiang nationality on the planning and design of the rural landscape of the Qiang nationality and proposes a strategy for the planning and design of the rural landscape of the Qiang nationality. This study hopes that this can provide theoretical and practical reference for the planning and design methods of the Qiang rural landscape in the upper reaches of the Minjiang River and the protection, development, and dissemination of the Qiang culture in the upper reaches of the Minjiang River. Experiments show that there are 764 settlements when the radius of the river buffer is within 500 meters. [ABSTRACT FROM AUTHOR]

In the current era of increasingly complex social life, as people's demand quality is getting higher and higher, the solution of the problem often has multiple indicators to achieve the optimum. This forces the graph coloring problem (GCP) to become complicated, and it is difficult to directly obtain the optimal solution, which brings new challenges to the solution of the problem. In response to this problem, the GCP field is of great research significance. With the in-depth study of GCP, the research on multiobjective optimization (MOO) in graph coloring algorithm (GCA) is gradually carried out. Its performance advantage is of great significance for solving multicondition constraint problems. This paper aims to study the application of Genetic Algorithm (GA) in GCA. Through the analysis and research of GA, and the fusion of numerical analysis and scientific computing, it can be applied to the construction of the neighbor distinguishable uniform V-full coloring algorithm (AVDEVTCA) to solve the AVDEVTC problem. This paper describes the basic theory of MOO and graph coloring. It conducts an experimental analysis of the algorithm performance and uses the relevant theoretical formulas to explain it. The results show that the algorithm takes 5754.142 S seconds to test 21325415 images and can color a large number of results that cannot be done manually in less time. It has greatly improved in terms of time and manpower saving and greatly improved practicability and work efficiency. [ABSTRACT FROM AUTHOR]

Block Gram-Schmidt algorithms serve as essential kernels in many scientific computing applications, but for many commonly used variants, a rigorous treatment of their stability properties remains open. This work provides a comprehensive categorization of block Gram-Schmidt algorithms, particularly those used in Krylov subspace methods to build orthonormal bases one block vector at a time. Known stability results are assembled, and new results are summarized or conjectured for important communication-reducing variants. Additionally, new block versions of low-synchronization variants are derived, and their efficacy and stability are demonstrated for a wide range of challenging examples. Numerical examples are computed with a versatile Matlab package hosted at https://github.com/katlund/BlockStab , and scripts for reproducing all results in the paper are provided. Block Gram-Schmidt implementations in popular software packages are discussed, along with a number of open problems. An appendix containing all algorithms type-set in a uniform fashion is provided. [ABSTRACT FROM AUTHOR]

Distributed-memory graph algorithms are fundamental enablers in scientific computing and analytics workflows. A majority of graph algorithms rely on the graph neighborhood communication pattern, i.e., repeated asynchronous communication between a vertex and its neighbors in the graph. The pattern is adversarial for communication software and hardware due to high message injection rates and input-dependent, many-to-one traffic with variable destinations and volumes. We present benchmarks and performance analysis of graph neighborhood communication on modern large-scale network interconnects from four supercomputers: ALCF Theta, NERSC Cori, OLCF Summit and R-CCS Fugaku. Our benchmarks characterize communication from the perspectives of latency and throughput. Benchmark parameters make it possible to mimic the behaviors of complex applications on real world or synthetic graphs by varying work distribution, remote edges, message volume, and per-vertex work. We find that minor changes in the input graph can substantially increase latencies; and contention can develop in memory caches and network stacks before contention in the network itself. Further, latencies and contention vary significantly for different graph neighborhoods, motivating the need for exploring asynchronous algorithms in greater detail. When adding work, load imbalance on real-world graphs can be pronounced: latencies for the 99th percentile were 8–128× than the corresponding average latencies. Our results help analysts and developers understand the performance implications of this important pattern, especially for the impending exascale platforms. [ABSTRACT FROM AUTHOR]

Deep learning (DL) has had unprecedented success and is now entering scientific computing with full force. However, current DL methods typically suffer from instability, even when universal approximation properties guarantee the existence of stable neural networks (NNs). We address this paradox by demonstrating basic well-conditioned problems in scientific computing where one can prove the existence of NNs with great approximation qualities; however, there does not exist any algorithm, even randomized, that can train (or compute) such a NN. For any positive integers K > 2 and L, there are cases where simultaneously 1) no randomized training algorithm can compute a NN correct to K digits with probability greater than 1/2; 2) there exists a deterministic training algorithm that computes a NN with K − 1 correct digits, but any such (even randomized) algorithm needs arbitrarily many training data; and 3) there exists a deterministic training algorithm that computes a NN with K − 2 correct digits using no more than L training samples. These results imply a classification theory describing conditions under which (stable) NNs with a given accuracy can be computed by an algorithm. We begin this theory by establishing sufficient conditions for the existence of algorithms that compute stable NNs in inverse problems. We introduce fast iterative restarted networks (FIRENETs), which we both prove and numerically verify are stable. Moreover, we prove that only O(| log(ϵ)|) layers are needed for an -accurate solution to the inverse problem. [ABSTRACT FROM AUTHOR]

The classification inaccuracy occurs to the evaluation calculation of the total factor production impact in the traditional digital economy development area. This paper applies the fuzzy hierarchical (Visualization in Scientific Computing, VISC) algorithm to the calculation method of the total factor production impact evaluation in the digital economy development area. The quantitative recursive method is used to evaluate the ability of the data information model, to achieve the ability of controlling the acquisition of characteristic resources, and complete the classification and summary of the index parameters of the total factor production impact of the digital economy development area. Finally, the experiment proves that this calculation method is used to develop the evaluation of the impact of total factor production digital economy development area and improve the information integration and analysis capabilities, the accuracy of the evaluation of the impact of total factor production, and the efficiency of the use of digital economy development resources. [ABSTRACT FROM AUTHOR]

Fortran is still widely used in scientific computing, and a very large corpus of legacy as well as new code is written in FORTRAN 77. In general this code is not type safe, so that incorrect programs can compile without errors. In this paper, we present a formal approach to ensure type safety of legacy Fortran code through automated program transformation. The objective of this work is to reduce programming errors by guaranteeing type safety. We present the first rigorous analysis of the type safety of FORTRAN 77 and the novel program transformation and type checking algorithms required to convert FORTRAN 77 subroutines and functions into pure, side-effect free subroutines and functions in Fortran 90. We have implemented these algorithms in a source-to-source compiler which type checks and automatically transforms the legacy code. We show that the resulting code is type safe and that the pure, side-effect free and referentially transparent subroutines can readily be offloaded to accelerators. [ABSTRACT FROM AUTHOR]

• Propose the aligned storage formats ACSR and AELL. • Put forward a parallel SpMV algorithm based on ACSR and AELL formats. • Theoretically analyze the performance of SpMV on ARM processors. • Evaluate the performance of our optimization on Kunpeng 920 processors. Sparse matrix-vector multiplication (SpMV) has always been a hot topic of research for scientific computing and big data processing, but the sparsity and discontinuity of the nonzero elements in a sparse matrix lead to the memory bottleneck of SpMV. In this paper, we propose aligned CSR (ACSR) and aligned ELL (AELL) formats and a parallel SpMV algorithm to utilize NEON SIMD registers on ARM processors. We analyze the impact of SIMD instruction latency, cache access, and cache misses on SpMV with different formats. In the experiments, our SpMV algorithm based on ACSR achieves 1.18x and 1.56x speedup over SpMV based on CSR and SpMV in PETSc, respectively, and AELL achieves 1.21x speedup over ELL. The deviations between the theoretical results and experimental results in the instruction latency and cache access are 10.26% and 10.51% in ACSR and 5.68% and 2.91% in AELL, respectively. [ABSTRACT FROM AUTHOR]

SpMV (Sparse matrix-vector multiplication) is an essential component in scientific computing and has attracted the attention of researchers in related fields at home and abroad. With the continuous expansion of matrix data, the efficient parallel SpMV algorithm has become a research hotspot for research experts in related fields. The sparse matrix compression format as a critical point to improve computing performance can effectively save storage space and efficiently cooperate with the advantages of the processor system structure to give full play to performance. This paper proposes a new sparse matrix storage format CSR2 (Compressed Sparse Row 2). It is a new single format and suitable for processor platforms with SIMD (Single Instruction Multiple Data) vectorizations. The format operation of CSR2 is easy to implement with a low overhead of conversion. We compared the SpMV algorithm based on CSR2 with the most advanced single format CSR5 (Compressed Sparse Row 5) and Intel MKL (Intel Math Kernel Library) on the mainstream high-performance processor Intel Xeon E5-2670 v3 CPU. We choose 48 sets of matrices to be used as a benchmark suite. Experimental results show that CSR2 has a remarkable performance improvement compared with CSR5 and MKL. Compared to CSR5, CSR2 can achieve an average acceleration of 1.401 × (up to 1.861 ×). Compared to MKL, CSR2 can achieve an average acceleration of 1.261 × (up to 5.921 ×). In reality, for applications with multiple iterations, using our CSR2 can bring low-overhead format conversion and high-throughput computing performance. [ABSTRACT FROM AUTHOR]

Least squares form one of the most prominent classes of optimization problems with numerous applications in scientific computing and data fitting. When such formulations aim at modeling complex systems, the optimization process must account for nonlinear dynamics by incorporating constraints. In addition, these systems often incorporate a large number of variables, which increases the difficulty of the problem and motivates the need for efficient algorithms amenable to large-scale implementations. In this paper, we propose and analyze a Levenberg--Marquardt algorithm for nonlinear least squares subject to nonlinear equality constraints. Our algorithm is based on inexact solves of linear least-squares problems that only require Jacobian-vector products. Global convergence is guaranteed by the combination of a composite step approach and a nonmonotone step acceptance rule. We illustrate the performance of our method on several test cases from data assimilation and inverse problems; our algorithm is able to reach the vicinity of a solution from an arbitrary starting point and can outperform the most natural alternatives for these classes of problems. [ABSTRACT FROM AUTHOR]

The efficient utilization of mixed-precision numerical linear algebra algorithms can offer attractive acceleration to scientific computing applications. Especially with the hardware integration of low-precision special-function units designed for machine learning applications, the traditional numerical algorithms community urgently needs to reconsider the floating point formats used in the distinct operations to efficiently leverage the available compute power. In this work, we provide a comprehensive survey of mixed-precision numerical linear algebra routines, including the underlying concepts, theoretical background, and experimental results for both dense and sparse linear algebra problems. [ABSTRACT FROM AUTHOR]

Computing reachability probabilities is a fundamental problem in the analysis of randomized programs. This article aims at a comprehensive and comparative account of various martingale-based methods for over- and under-approximating reachability probabilities. Based on the existingworks that stretch across different communities (formal verification, control theory, etc.), we offer a unifying account. In particular, we emphasize the role of order-theoretic fixed points--a classic topic in computer science--in the analysis of randomized programs. This leads us to two new martingale-based techniques, too. We also make an experimental comparison using our implementation of template-based synthesis algorithms for those martingales. [ABSTRACT FROM AUTHOR]

We propose a multi-level method to increase the accuracy of machine learning algorithms for approximating observables in scientific computing, particularly those that arise in systems modelled by differential equations. The algorithm relies on judiciously combining a large number of computationally cheap training data on coarse resolutions with a few expensive training samples on fine grid resolutions. Theoretical arguments for lowering the generalisation error, based on reducing the variance of the underlying maps, are provided and numerical evidence, indicating significant gains over underlying single-level machine learning algorithms, are presented. Moreover, we also apply the multi-level algorithm in the context of forward uncertainty quantification and observe a considerable speedup over competing algorithms. [ABSTRACT FROM AUTHOR]

Sparse triangular solve (SpTRSV) is an important kernel in many scientific computing applications. In traditional viewpoints, accelerating SpTRSV by parallelizing the solution process is a challenging task. Dependencies among the variables that exist in the solution process not only restrict the parallelism that can be achieved, but also introduce large synchronization overhead among the parallel tasks. Moreover, a time-consuming pre-processing phase is commonly required to identify calculations that can be parallelized. However, we have observed that a large number of dependencies among the variables can be eliminated if we only calculate partial values of the variables first and add them together to obtain the final values later. By using such a strategy, starting to solve a variable does not need to wait for all of its prerequisite variables having been solved. In consequence, parallelism of the SpTRSV can be increased significantly. In this paper, we transform above mentioned observations into a subtree-based parallel algorithm to accelerate SpTRSV. The proposed algorithm calculates partial values of the variable along with an implicit subtree traversal and utilizes hardware atomic operation to implement accumulation of the partial values. This not only introduces no pre-processing overhead, but also avoids any barrier synchronization among the parallel threads. We evaluate the proposed algorithm on 2135 matrices from SuiteSparse Matrix Collection based on a generic GPU platform. Experimental results demonstrate that our scheme outperforms the state-of-the-art GPU and CPU vendor libraries in 1949 and 1782 matrices, respectively. Compared with the latest synchronization-free method, our scheme outperforms in 1779 matrices. [ABSTRACT FROM AUTHOR]

We propose a supervised deep learning algorithm based on low-discrepancy sequences as the training set. By a combination of theoretical arguments and extensive numerical experiments we demonstrate that the proposed algorithm significantly outperforms standard deep learning algorithms that are based on randomly chosen training data for problems in moderately high dimensions. The proposed algorithm provides an efficient method for building inexpensive surrogates for many underlying maps in the context of scientific computing. [ABSTRACT FROM AUTHOR]

In this work, we present an adaptive unfitted finite element scheme that combines the aggregated finite element method with parallel adaptive mesh refinement. We introduce a novel scalable distributed-memory implementation of the resulting scheme on locally adapted Cartesian forest-of-trees meshes. We propose a two-step algorithm to construct the finite element space at hand by means of a discrete extension operator that carefully mixes aggregation constraints of problematic degrees of freedom, which get rid of the small cut cell problem, and standard hanging degree of freedom constraints, which ensure trace continuity on nonconforming meshes. Following this approach, we derive a finite element space that can be expressed as the original one plus well-defined linear constraints. Moreover, it requires minimum parallelization effort, using standard functionality available in existing large-scale finite element codes. Numerical experiments demonstrate its optimal mesh adaptation capability, robustness to cut location, and parallel efficiency, on classical Poisson hp-adaptivity benchmarks. Our work opens the path to functional and geometrical error-driven dynamic mesh adaptation with the aggregated finite element method in large-scale realistic scenarios. Likewise, it can offer guidance for bridging other scalable unfitted methods and parallel adaptive mesh refinement. [ABSTRACT FROM AUTHOR]

• Crack tracing algorithm can be combined with a variety of crack-bridge models. • Cracks are implicitly detected on a one-by-one basis. • Implementation using open-source scientific computing tools with a web application. The paper describes a probabilistic model capturing the multiple-cracking behavior of unidirectional brittle-matrix composites loaded in tension. The approach to modeling of composite fragmentation introduces two salient features that enhance both efficiency and flexibility compared to existing simulation methods: First, the algorithm identifies the emerging cracks one by one within a minimum number of load increments. Second, the crack-tracing algorithm is based on an abstract description of a crack bridge behavior. As a result, it is possible to combine the crack-tracing algorithm with a wide variety of crack bridge models, i.e. non-linear, deterministic or probabilistic. In this way, specific phenomena of bond behavior in different types of composites can be accounted for. The random nature of matrix cracking is reproduced using a random field simulation of the matrix strength. The model is verified by reproducing analytical results available for constant bond-slip law. The feasibility and robustness of the algorithm is demonstrated using an interactive web application that is directly executable from within a public github.com repository. [ABSTRACT FROM AUTHOR]

This paper proposes a standard documentary framework called Databook for Data Science projects. This proposal is the result of five years of action-research on multiple projects in several sectors of activity in France and a confrontation of standard theoretical processes of Data Science, such as CRISP_DM, with the reality of the field. The minimalist and flexible structure of the Databook prototype, described and illustrated in this paper, has revealed its operationality on more than a hundred projects and has been recognised by various stakeholders as an excellent facilitator of Human Data Mediation, especially for multi-skilled projects. Beyond its proven benefits for project efficiency, this framework, conceived as a frontier object, can be applied more broadly to data project portfolio management and data value, governance and quality. By surpassing the computational aspect of the models, the Databook is an answer to the issues of interpretability and auditability of algorithms. [ABSTRACT FROM AUTHOR]

Rendezvous algorithms encode a communication pattern that is useful when processors sending data do not know who the receiving processors should be, or vice versa. The idea is to define an intermediate decomposition where datums from different sending processors can "rendezvous" to perform a computation, in a manner that both the senders and eventual receivers of the results can identify the appropriate rendezvous processor. Originally designed for interpolating between overlaid grids with independent parallel decompositions (Plimpton et al., 2004), we have recently found rendezvous algorithms useful for a variety of operations in particle- or grid-based simulation codes when running large problems on large numbers of processors. In particular, we show they can perform well when a load-balanced intermediate decomposition is randomized and not spatial, requiring all-to-all communication to move data between processors. In this case rendezvous algorithms leverage the large bisection communication bandwidths which parallel machines provide. We describe how rendezvous algorithms work in a scientific computing context and give specific examples for molecular dynamics and Direct Simulation Monte Carlo codes which result in dramatic performance improvements versus simpler algorithms which do not scale as well. We explain how a generic rendezvous algorithm can be implemented, and also point out similarities with the MapReduce paradigm popularized by Google and Hadoop. • This paper explains how parallel rendezvous algorithms work in the context of large-scale scientific modeling and simulation codes. • It gives several examples of such algorithms. • It presents performance gains of the new algorithms relative to previously used simpler algorithms which were less scalable. [ABSTRACT FROM AUTHOR]

On the basis of studying datasets of students' course scores, we constructed a Bayesian network and undertook probabilistic inference analysis. We selected six requisite courses in computer science as Bayesian network nodes. We determined the order of the nodes based on expert knowledge. Using 356 datasets, the K2 algorithm learned the Bayesian network structure. Then, we used maximum a posteriori probability estimation to learn the parameters. After constructing the Bayesian network, we used the message-passing algorithm to predict and infer the results. Finally, the results of dynamic knowledge inference were presented through a detailed inference process. In the absence of any evidence node information, the probability of passing other courses was calculated. A mathematics course (a basic professional course) was chosen as the evidence node to dynamically infer the probability of passing other courses. Over time, the probability of passing other courses greatly improved, and the inference results were consistent with the actual values and can thus be visualized and applied to an actual school management system. [ABSTRACT FROM AUTHOR]

This paper addresses the problem of designing a β -additive fault-tolerant approximate BFS (or FT-ABFS for short) structure, namely, a subgraph H of the network G such that subsequent to the failure of a single edge e, the surviving part of H still contains an approximate BFS spanning tree for (the surviving part of) G, whose distances satisfy dist (s , v , H \ { e }) ≤ dist (s , v , G \ { e }) + β for every v ∈ V . It was shown in Parter and Peleg (SODA, 2014), that for every β ∈ [ 1 , O (log n) ] there exists an n-vertex graph G with a source s for which any β -additive FT-ABFS structure rooted at s has Ω (n 1 + ϵ (β)) edges, for some function ϵ (β) ∈ (0 , 1) . In particular, 3-additive FT-ABFS structures admit a lower bound of Ω (n 5 / 4) edges. In this paper we present the first upper bound, showing that there exists a poly-time algorithm that for every n-vertex unweighted undirected graph G and source s constructs a 4-additive FT-ABFS structure rooted at s with at most O (n 4 / 3) edges. The main technical contribution of our algorithm is in adapting the path-buying strategy used in Baswana et al. (ACM Trans Algorithms 7:A5, 2010) and Cygan et al. (Proceedings of the 30th symposium on theoretical aspects of computer science, pp 209–220, 2013) to failure-prone settings. [ABSTRACT FROM AUTHOR]

Mixed-cell-height circuits have become popular in advanced technologies for better power, area, routability, and performance tradeoffs. With technology and region constraints imposed by modern circuit designs, the mixed-cell-height legalization problem has become even more challenging. Additionally, an ideal legalization method should minimize both the average and maximum cell movements to preserve the quality of a given placement as much as possible. In this article, we present an effective and efficient mixed-cell-height legalization algorithm to consider technology and region constraints while minimizing the average and maximum cell movements. We first present a fence region handling technique to unify the fence regions and the default region. To obtain a desired cell assignment, we then propose a movement-aware cell reassignment method by iteratively reassigning cells in locally dense areas to their desired rows. After cell reassignment, a technology-aware legalization is presented to remove cell overlaps while satisfying the technology constraints. Finally, we propose a technology-aware refinement to further reduce the average and maximum cell movements without increasing the technology constraints violations. Compared with the champion of the 2017 CAD Contest at ICCAD and the state-of-the-art work, experimental results based on the 2017 CAD Contest at ICCAD benchmarks show that our algorithm achieves the best average and maximum cell movements and significantly fewer technology constraint violations, in a comparable runtime. The experimental results based on the modified 2015 ISPD Contest benchmarks also demonstrate the effectiveness of our algorithm in minimizing the average and maximum cell movements, compared with state-of-the-art mixed-cell-height legalizers. [ABSTRACT FROM AUTHOR]

Noise and missing data are intrinsic characteristics of real-world data, leading to uncertainty that negatively affects the quality of knowledge extracted from the data. The burden imposed by missing data is often severe in sensors that collect data from the physical world, where large gaps of missing data may occur when the system is temporarily off or disconnected. How can we reconstruct missing data for these periods? We introduce an accurate and efficient algorithm for missing data reconstruction (imputation), that is specifically designed to recover off-period segments of missing data. This algorithm, Ghost, searches the sequential dataset to find data segments that have a prior and posterior segment that matches those of the missing data. If there is a similar segment that also satisfies the constraint – such as location or time of day – then it is substituted for the missing data. A baseline approach results in quadratic computational complexity, therefore we introduce a caching approach that reduces the search space and improves the computational complexity to linear in the common case. Experimental evaluations on five real-world datasets show that our algorithm significantly outperforms four state-of-the-art algorithms with an average of 18 percent higher F-score. [ABSTRACT FROM AUTHOR]

In the problem of compressed phase retrieval, the goal is to reconstruct a sparse or approximately $k$ -sparse vector $x \in \mathbb {C} ^{n}$ given access to $y= |\Phi x|$ , where $|v|$ denotes the vector obtained from taking the absolute value of $v\in \mathbb {C} ^{n}$ coordinate-wise. In this paper we present sublinear-time algorithms for a few for-each variants of the compressive phase retrieval problem which are akin to the variants considered for the classical compressive sensing problem in theoretical computer science. Our algorithms use pure combinatorial techniques and near-optimal number of measurements. [ABSTRACT FROM AUTHOR]

The research focus of an institution indicates the academic strength of the institution. This academic strength goes a long way to determine the quality of academic product expected from such an institution. To this effect, this study aims at analyzing the research publication strength of 70 Nigerian universities that are visible in Scopus database for Computer Science from inception to 22 December 2017. They include 35 federal, 27 state and 8 private universities. This study is divided into two phases; the first phase consists of six major steps in a waterfall model and the second is carried out by adopting a knowledge discovery model using a clustering algorithm for pattern discovery in an unstructured data model. The data source for the analysis is the Scopus indexing database recommended by the Times Higher Education for international ranking of higher institutions. From the first phase, the predominant outlet for each institution was discovered alongside the percentage of publications in the dominant outlet and the number of publications by the institution within the database over the period of the analysis. In the second phase, the predominant areas of research of each institution within the field of Computer Science was determined. In particular, it was observed that such areas as deep learning and data science are under researched. The findings also revealed networking as the most published area in Nigerian universities, and a number of Nigerian computer science scholars engage in application research which link Computer Science with other disciplines with Nigeria and Africa as case studies. The findings from this research can support postgraduate students' decision in making the right choice of institution based on their research interest. It will also serve as an eye-opener for industries in promoting a result-oriented university-industry partnership. [ABSTRACT FROM AUTHOR]

The Residue Logarithmic Number System (RLNS) in digital mathematics allows multiplication and division to be performed considerably quickly and more precisely than the extensively used Floating-Point number setups. RLNS in the pitch of large scale integrated circuits, digital signal processing, multimedia, scientific computing and artificial neural network applications have Fixed Width property which has equal number of in and out bit width; hence, these applications need a Fixed Width multiplier. In this paper, a Fixed Width-Floating-Point multiplier based on RLNS was proposed to increase the processing speed. The truncation errors were reduced by using Taylor series. RLNS is the combination of both the residue number system and the logarithmic number system, and uses a table lookup including all bits for expansion. The proposed scheme is effective with regard to speed, area and power utilization in contrast to the design of conservative Floating-Point mathematics designs. Synthesis results were obtained using a Xilinx 14.7 ISE simulator. The area is 16,668 µm2, power is 37 mW, delay is 6.160 ns and truncation error can be lessened by 89% as compared with the direct-truncated multiplier. The proposed Fixed Width RLNS multiplier performs with lesser compensation error and with minimal hardware complexity, particularly as multiplier input bits increment. [ABSTRACT FROM AUTHOR]

Finite topological spaces and their dimensions have many applications in computer science, e.g., in digital topology, computer graphics and the analysis and synthesis of digital images. Georgiou et. al. [11] provided a polynomial algorithm for computing the covering dimension dim(X;) of a finite topological space (X;). In addition, they asked whether algorithms of the same complexity for computing the small inductive dimension ind(X;) and the large inductive dimension Ind(X;) can be developed. The first problem was solved in a previous paper [4]. Using results of the that paper, we also solve the second problem in this paper. We present a polynomial algorithm for Ind(X;), so that there are now efficient algorithms for the three most important notions of a dimension in topology. Our solution reduces the computation of Ind(X;), where the specialisation pre-order of (X;) is taken as input, to the computation of the maximal height of a specific class of directed binary trees within the partially ordered set. For the latter an efficient algorithm is presented that is based on order- and graph-theoretic ideas. Also refinements and variants of the algorithm are discussed. [ABSTRACT FROM AUTHOR]

"What is an algorithm?" is a fundamental question of computer science. Gurevich's behavioural theory of sequential algorithms (aka the sequential ASM thesis) gives a partial answer by defining (non-deterministic) sequential algorithms axiomatically, without referring to a particular machine model or programming language, and showing that they are captured by (nondeterministic) sequential Abstract State Machines (nd-seq ASMs). However, recursive algorithms such as mergesort are not covered by this theory, as has been pointed out by Moschovakis, who had independently developed a different framework to mathematically characterize the concept of (in particular recursive) algorithm. In this article we propose an axiomatic definition of the notion of sequential recursive algorithm which extends Gurevich's axioms for sequential algorithms by a Recursion Postulate and allows us to prove that sequential recursive algorithms are captured by recursive Abstract State Machines, an extension of nd-seq ASMs by a CALL rule. Applying this recursive ASM thesis yields a characterization of sequential recursive algorithms as finitely composed concurrent algorithms all of whose concurrent runs are partial-order runs. [ABSTRACT FROM AUTHOR]

In the first part of our article we will refer the penetration of scientific terms into colloquial language, focusing on the sense in which the concept of an algorithm currently functions outside its original scope. The given examples will refer mostly to disciplines not directly related to computer science and to the colloquial language. In the next part we will also discuss the modifications made to the meaning of the term algorithm and how this concept is now understood in computer science. Finally, we will discuss the problem of algorithmicity of evolutionary algorithms, i.e. we will try to answer the question whether - and to what extent - they are still algorithms in the classical sense. [ABSTRACT FROM AUTHOR]

Given a graph G, a properk-coloring of G is a partition c = (S i) i ∈ [ 1 , k ] of V(G) into k stable sets S 1 , ... , S k . Given a weight function w : V (G) → R + , the weight of a color S i is defined as w (i) = max v ∈ S i w (v) and the weight of a coloringc as w (c) = ∑ i = 1 k w (i) . Guan and Zhu (Inf Process Lett 61(2):77–81, 1997) defined the weighted chromatic number of a pair (G, w), denoted by σ (G , w) , as the minimum weight of a proper coloring of G. The problem of determining σ (G , w) has received considerable attention during the last years, and has been proved to be notoriously hard: for instance, it is NP-hard on split graphs, unsolvable on n-vertex trees in time n o (log n) unless the ETH fails, and W[1]-hard on forests parameterized by the size of a largest tree. We focus on the so-called dual parameterization of the problem: given a vertex-weighted graph (G, w) and an integer k, is σ (G , w) ≤ ∑ v ∈ V (G) w (v) - k ? This parameterization has been recently considered by Escoffier (in: Proceedings of the 42nd international workshop on graph-theoretic concepts in computer science (WG). LNCS, vol 9941, pp 50–61, 2016), who provided an FPT algorithm running in time 2 O (k log k) · n O (1) , and asked which kernel size can be achieved for the problem. We provide an FPT algorithm in time 9 k · n O (1) , and prove that no algorithm in time 2 o (k) · n O (1) exists under the ETH. On the other hand, we present a kernel with at most (2 k - 1 + 1) (k - 1) vertices, and rule out the existence of polynomial kernels unless NP ⊆ coNP / poly , even on split graphs with only two different weights. Finally, we identify classes of graphs allowing for polynomial kernels, namely interval graphs, comparability graphs, and subclasses of circular-arc and split graphs, and in the latter case we present lower bounds on the degrees of the polynomials. [ABSTRACT FROM AUTHOR]

Digital Bibliography and Library Project dataset is a collection of bibliographic records of computer science publications of various authors and co-authors. It contains approximately 1.5 million bibliographic records. An algorithm for an author's information retrieval is developed to retrieve details of specific author publications and correlation among authors. Further performance of an author is measured with parameters like consistency, contribution factor, stability, cooperativeness, and solidity. The work presented is tested on the DBLP dataset. Experimental results clearly support the claim that it works efficiently for retrieving specific author-publication records and its analysis with respect to suggested parameters. [ABSTRACT FROM AUTHOR]

Today, Sentinel ‐ 1A data with terrain observation with progressive scanning (TOPS) imaging mode is increasingly used in earth observation aiming at consistent monitoring of surface change and its deformation. However, due to the limited accuracy of coarse co ‐registration, spectral aliasing along the azimuth direction enables the presence of phase jumping in overlapping area of neighboring bursts. Although geometrical co‐registration in conjunction with enhanced spectral diversity (ESD) is proven to be a feasible strategy to correct such error and has been widely used in some open source softwares (e.g. DORIS (Delft object‐oriented radar interferometric software), SNAP (sentinel applications platform), ISCE (InSAR scientific computing environment)), the performance of ESD is still not satisfactory and relies strongly on spatiotemporal decorrelation. Given that decorrelation is generally quantized by interferometric coherence, this paper presents and assesses a new methodology to improve the accuracy of ESD by fully exploring the high coherent targets. Specifically, this method focuses on mitigating the spatiotemporal decorrelation in fine coregistration procedures by: (1) Selecting stable targets with moderate and high coherence using accurate coherence estimation. (2) Maximizing coherence magnitude by optimal interferometric subset chosen from Dijkstra algorithm. We compare and test this method against current favorites based on single master image and network‐based enhanced spectral diversity (NESD), and the experimental results demonstrate the value of our method. It can make up for the shortage of NESD method. [ABSTRACT FROM AUTHOR]

In this paper, we develop a novel spectral method for solving the time-dependent Kohn{Sham equations in the semiclassical regime. Our strategy is to use a single-step predictor- corrector algorithm, where the propagator is derived using a Fourier integral operator commonly known as the frozen Gaussian approximation (FGA) ansatz. In the case of laser potentials, we derived a simplified FGA to avoid the high dimensional integration. Numerical examples are provided to verify applicability, as well as efficiency and accuracy of the scheme. [ABSTRACT FROM AUTHOR]

The time complexity of algorithms, i.e., the scaling of the time required for solving a problem instance as a function of instance size, is of key interest in theoretical computer science and practical applications. In this work, we present a fully automated tool – Empirical Scaling Analyzer (ESA) – for performing sophisticated and detailed empirical scaling analyses. The methodological approach underlying ESA is based on a combination of automatic function fitting and bootstrap sampling; previous versions of the methodology have been used in prior work to characterize the empirical scaling behaviour of several prominent, high-performance SAT and TSP solvers. ESA is applicable to any algorithm or system, as long as running time data can be collected on sets of problem instances of various sizes. We present results from rigorous stress-testing to critically assess ESA on scenarios with challenging characteristics. We also give an overview of empirical scaling results obtained using ESA. [ABSTRACT FROM AUTHOR]

Set functions with convenient properties (such as submodularity) appear in application areas of current interest, such as algorithmic game theory, and allow for improved optimization algorithms. It is natural to ask (e.g., in the context of data driven optimization) how robust such properties are, and whether small deviations from them can be tolerated. We consider two such questions in the important special case of linear set functions. One question that we address is whether any set function that approximately satisfies the modularity equation (linear functions satisfy the modularity equation exactly) is close to a linear function. The answer to this is positive (in a precise formal sense) as shown by Kalton and Roberts [Trans. Amer. Math. Soc., 278 (1983), pp. 803--816] (and further improved by Bondarenko, Prymak, and Radchenko [J. Math. Anal. Appl., 402 (2013), pp. 234--241]). We revisit their proof idea that is based on expander graphs and provide significantly stronger upper bounds by combining it with new techniques. Furthermore, we provide improved lower bounds for this problem. Another question that we address is that of how to learn a linear function h that is close to an approximately linear function f, while querying the value of f on only a small number of sets. We present a deterministic algorithm that makes only linearly many (in the number of items) nonadaptive queries, and thus improve upon a previous algorithm of Chierichetti, Das, Dasgupta, and Kumar [Proceedings of the 56th Symposium on Foundations of Computer Science, 2015, pp. 1143--1162] that is randomized and makes more than a quadratic number of queries. Our learning algorithm is based on the Hadamard transform. [ABSTRACT FROM AUTHOR]

Aiming at the problems of slow convergence speed, easy to fall into local optimum and low solution quality of carnivorous plant algorithm (CPA) in the existing literature, a carnivorous plant algorithm with Lévy mutation and similarity-removal operation (CPA-LMSR) is proposed. In the grouping stage, a method of grouping according to the biological species, carnivorous plants and prey, is proposed to enhance the randomness of sampling. In the growth stage, the adaptive variation of attraction rate and growth rate parameters are given. An attraction term is added to the growth model of carnivorous plants and prey, which improves the probability of generating potential offspring. During the reproduction phase, all carnivorous plants are allowed to reproduce, ensuring excellent individual information to be retained with greater probability. A similarity-removal operation and Lévy mutation operator are added to improve the exploration ability of the algorithm to maintain the diversity of the population. Finally, CPA-LMSR is tested on 28 constrained optimization problems (COPs) in CEC 2017 and two practical engineering optimization problems, and compared with seven comparison algorithms in the literature. The experimental results show that CPA-LMSR has better solution quality, stability and convergence speed than other algorithms, and the performance of various algorithms is significantly different, which is proved that CPA-LMSR can effectively solve complex constrained optimization models and provide efficient computing power for scientific decision-making in various fields of science, engineering and economy, so as to solve optimization problems in various fields. [ABSTRACT FROM AUTHOR]

Abstract A modified multilevel algorithm for solving the excessive storage requirements and ill-conditioning encountered in the boundary-type discretization method is proposed. The modified multilevel algorithm is an extension of the modified dual-level algorithm from dual levels to multiple levels. The method is a kernel-independent method. The core idea is the layer-by-layer calculation and then layer-by-layer correction. Making use of a multilevel structure, the original sparse matrix of the modified dual-level algorithm breaks down into a series of smaller sparse matrices corresponding to different fine meshes. The final matrix to be solved is hereby transformed to a series of smaller sparse matrices instead of a fully-populated matrix. The preconditioning effect originating from the recursive computations among the coarse mesh and fine meshes constitutes its core competitive attribute. The method evaluates far-field contributions only by the coarse mesh and uses a gradual approach to evaluate the near-field contributions. The storage requirements and computing complexity are hereby further reduced significantly. [ABSTRACT FROM AUTHOR]

Since the 1950s and with the emergence of computers and Computer Science new perspectives opened up, new ways of thinking that profoundly changed the world. This article seeks to understand this new way of thinking and, although the philosophy of science deeply analyzes the traditional way of scientific thinking, Computational Sciences are still an unknown terrain for it. Because it is difficult to ask the right questions when you do not have concepts, theories or frameworks to formulate them. In addition, while people believe that they are aware of and properly understand the object of most sciences, not all are well understood. For example, although Nature Sciences study the world to develop new knowledge about the universe, what new ideas can be found in a device made by man as the computer? For that reason, some affirm that the Computer Sciences are an abstract field, as the mathematics, and that they do not really treat of computers, but of another thing. That is why they are still a scientific discipline that is still a source of much confusion. This article aims to develop an understanding of Computer Sciences. [ABSTRACT FROM AUTHOR]

A large number of free boundary problems can be formulated as linear-complementarity problems. In this paper, we propose an inexact alternating direction method of multipliers for solving linear complementarity problem arising from free boundary problems by using the special structure of these problems. The convergence of our proposed method is proved. Numerical results show that the proposed method is feasible and effective, and it is significantly faster than modified alternating direction implicit algorithm and many other methods, especially when dimension of the problem being solved is large. [ABSTRACT FROM AUTHOR]

In the past decades, simulation frameworks have greatly increased in complexity, due to coupling of models from various disciplines into so-called integrated models. Recently, the combination with tools for uncertainty quantification, inverse modelling, optimization and control started a development towards what we call extended simulation frameworks. While there is an ongoing discussion on quality assurance and reproducibility for simulation frameworks, we have not observed a similar discussion for the extended case. Particularly for extended frameworks, the need for quality assurance is high: The overwhelming range of options and algorithms is unmanageable by a domain expert and opaque to decision makers or the public. The resulting demand for ‘intelligent software’ with automated configuration can lead to a blind trust in simulation results even if they are incorrect. This is a threatening scenario due to potential consequences in simulation-based engineering or political decisions. In this paper, we analyze the increasing complexity of scientific computing workflows, and discuss the corresponding problems of extended scientific simulation frameworks. We propose a paradigm that regulates the allowable properties of framework components, supports the framework configuration for complex simulations, enforces automatic self-tests of configured frameworks, and communicates automated algorithm choices, potentially critical user settings or convergence issues with adaptive detail level and urgency to the end-user. Our goal is to start transferring the quality assurance discussion in the field of integrated modeling and conventional software frameworks to the area of extended simulation frameworks. With this, we hope to increase the reliability and transparency of (extended) frameworks, framework use and of the corresponding simulation results. [ABSTRACT FROM AUTHOR]

Reduced basis methods are projection-based model order reduction techniques for reducing the computational complexity of solving parametrized partial differential equation problems. In this work we discuss the design of pyMOR, a freely available software library of model order reduction algorithms, in particular reduced basis methods, implemented with the Python programming language. As its main design feature, all reduction algorithms in pyMOR are implemented generically via operations on well-defined vector array, operator, and discretization interface classes. This allows for an easy integration with existing open-source high-performance partial differential equation solvers without adding any model reduction specific code to these solvers. Besides an in-depth discussion of pyMOR's design philosophy and architecture, we present several benchmark results and numerical examples showing the feasibility of our approach. [ABSTRACT FROM AUTHOR]

A kinetic Monte Carlo model of a reversible addition-fragmentation chain transfer (RAFT) process is presented. The algorithm has been developed and implemented in Julia for the three main RAFT theories under current discussion (slow fragmentation, intermediate radical termination, and intermediate radical termination with oligomers). Julia is a modern programming language designed to achieve high performance in numerical and scientific computing. Thanks to a careful optimization of the code, it is possible to simulate a RAFT reaction scheme in short computing times for any of the three theories. The code is benchmarked against other programming languages (MATLAB, Python, FORTRAN, and C), showing that Julia presents advantages for this particular system. The model offers an efficient method for predicting average properties and molecular weight distributions of the polymer species, including the bivariate molecular weight distribution of the intermediate two-arm adduct. The proposed model can also be employed to obtain additional detailed information regarding the polymer microstructure at any reaction time. [ABSTRACT FROM AUTHOR]

In the past 10 years, the 'parareal' (parallel-in-time) algorithm has attracted lots of attention thanks to its excellent performance in scientific computing. The parareal algorithm is iterative and is characterized by two propagators G and F which are associated with a coarse step size ΔT and a fine step size Δt, respectively, where ΔT =JΔt and J ≥ 2 is an integer. When we apply this algorithm to large-scale systems of ordinary differential equations obtained by semidiscretizing partial differential equations, two questions arise naturally. (I) Is the error between the iterate and the target solution contractive at each iteration for any choice of the discretization parameters ΔT, J and Δx? (II) How small can the contraction factor be and can such a contraction factor be independent of the discretization parameters? For linear problems u' =Au + g with symmetric negative-definite matrix A, when the implicit Euler method is used as both the G- and F-propagators, positive answers to these two questions were given by Mathew et al. (2010, SIAM J. Sci. Comput., 32, 1180-1200) and the contraction factor can be bounded by 0.298 for any choice of the discretization parameters. In this paper, for the case that the implicit Euler method is used as the G-propagator, we provide a positive answer to (I) for three second-order Fpropagators: the trapezoidal method, the TR/BDF2 method and the two-stage diagonally implicit Runge- Kutta (2s-DIRK) method. For (II), we prove that the contraction factors can be bounded by 0.316 and 13 for the 2s-DIRK method and the TR/BDF2 method (provided the parameter γ involved in TR/BDF2 satisfies γ ∈ [0.043, 0.977]), respectively, and both bounds are independent of the discretization parameters. For the trapezoidal method, we show that a uniform bound (less than 1) of the contraction factor does not exist. Numerical results are presented to validate the theoretical prediction. [ABSTRACT FROM AUTHOR]

High-performance computing is important in many areas of engineering. Increasing capabilities of modern workstations open new possibilities in scientific computing. This paper demonstrates how graphical processing units can be used efficiently for large models of elastic wave propagation in complex media. The method is based on the local interaction simulation approach and a parallel algorithm architecture. The focus of the work presented is on numerical implementation and covers aspects related to software modular architecture, computer memory organisation and optimisation. A domain decomposition approach allowing for calculations using multiple-GPU configurations is proposed, implemented and examined. The performance of the proposed simulation framework is tested for numerical models of different sizes, various computing precisions and hardware platforms. The results obtained are discussed in terms of graphical processing unit limitations. Obtained results indicate significant speed-up factors comparing to calculations using central processing units or different modelling approaches. [ABSTRACT FROM AUTHOR]