Researcher profile

Leonid B. Sokolinsky

Leonid B. Sokolinsky contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate

math.OC Distributed, Parallel, and Cluster Computing math.NA Mathematical Software Numerical Analysis Performance

Trust snapshot

Quick read

Trust 17 - UnverifiedVerification L1Unclaimed author

4works

0followers

6topics

2close collaborators

Actions

Decide how to stay connected

Follow researcher0

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2022arXiv

Visualizing Multidimensional Linear Programming Problems

The article proposes an n-dimensional mathematical model of the visual representation of a linear programming problem. This model makes it possible to use artificial neural networks to solve multidimensional linear optimization problems, the feasible region of which is a bounded non-empty set. To visualize the linear programming problem, an objective hyperplane is introduced, the orientation of which is determined by the gradient of the linear objective function: the gradient is the normal to the objective hyperplane. In the case of searching a maximum, the objective hyperplane is positioned in such a way that the value of the objective function at all its points exceeds the value of the objective function at all points of the feasible region, which is a bounded convex polytope. For an arbitrary point of the objective hyperplane, the objective projection onto the polytope is determined: the closer the objective projection point is to the objective hyperplane, the greater the value of the objective function at this point. Based on the objective hyperplane, a finite regular set of points is constructed, called the receptive field. Using objective projections, an image of a polytope is constructed. This image includes the distances from the receptive points to the corresponding points of the polytope surface. Based on the proposed model, parallel algorithms for visualizing a linear programming problem are constructed. An analytical estimation of its scalability is performed. Information about the software implementation and the results of large-scale computational experiments confirming the efficiency of the proposed approaches are presented.

preprint2021arXiv

BSF: a parallel computation model for scalability estimation of iterative numerical algorithms on cluster computing systems

This paper examines a new parallel computation model called bulk synchronous farm (BSF) that focuses on estimating the scalability of compute-intensive iterative algorithms aimed at cluster computing systems. In the BSF model, a computer is a set of processor nodes connected by a network and organized according to the master/slave paradigm. A cost metric of the BSF model is presented. This cost metric requires the algorithm to be represented in the form of operations on lists. This allows us to derive an equation that predicts the scalability boundary of a parallel program: the maximum number of processor nodes after which the speedup begins to decrease. The paper includes several examples of applying the BSF model to designing and analyzing parallel nu-merical algorithms. The large-scale computational experiments conducted on a cluster computing system confirm the adequacy of the analytical estimations obtained using the BSF model.

preprint2020arXiv

Scalable Method for Linear Optimization of Industrial Processes

In the development of industrial digital twins, the optimization problem of technological and business processes often arises. In many cases, this problem can be reduced to a large-scale linear programming (LP) problem. The article is devoted to the new method for solving large-scale LP problems. This method is called the "apex-method". The apex-method uses the predictor-corrector framework. The predictor step calculates a point belonging to the feasible region of LP problem. The corrector step calculates a sequence of points converging to the exact solution of the LP problem. The article gives a formal description of the apex-method and provides information about its parallel implementation in C++ language by using the MPI library. The results of large-scale computational experiments on a cluster computing system to study the scalability of the apex method are presented.

preprint2020arXiv

Scalable parallel algorithm for solving non-stationary systems of linear inequalities

In this paper, a scalable iterative projection-type algorithm for solving non-stationary systems of linear inequalities is considered. A non-stationary system is understood as a large-scale system of inequalities in which coefficients and constant terms can change during the calculation process. The proposed parallel algorithm uses the concept of pseudo-projection which generalizes the notion of orthogonal projection. The parallel pseudo-projection algorithm is implemented using the parallel BSF-skeleton. An analytical estimation of the algorithm scalability boundary is obtained on the base of the BSF cost metric. The large-scale computational experiments were performed on a cluster computing system. The obtained results confirm the efficiency of the proposed approach.

Institution

Affiliation not imported yet

This author record came from a source that does not expose affiliation metadata. Once the author claims the profile or we enrich the record from another provider, this section will link to the concrete institution.

Topic footprint