Who am I?
I’m a PhD student in Computer Science at Clemson University. I work under the supervision of Ilya Safro in the Algorithms and Computational Science Lab.
Before that, I’ve received a Bachelors degree in Applied Mathematics and Physics from Moscow Institute of Physics and Technology.
Here’s my CV.
What do I do?
My research focusses around hypergraph algorithms, multiscale methods, combinatorial optimization and quantum computing.
Check out my awesome projects!
How can I be reached?
Through any of the following links:
Aggregative Coarsening for Multilevel Hypergraph Partitioning
Algorithms for many hypergraph problems, including partitioning, utilize multilevel frameworks to achieve a good trade-off between the performance and the quality of results. In this paper we introduce two novel aggregative coarsening schemes and incorporate them within state-of-the-art hypergraph partitioner Zoltan. Our coarsening schemes are inspired by the algebraic multigrid and stable matching approaches. We demonstrate the effectiveness of the developed schemes as a part of multilevel hypergraph partitioning framework on a wide range of problems.
Available in PDF on arXiv.org
Relaxation-Based Coarsening for Multilevel Hypergraph Partitioning
Submitted: Fall 2017
Multilevel hypergraph partitioning, as well as many other algorithms on hypergraphs, relies on a measure of similarity between vertices to make decisions. In hypergraph partitioning setting, the decision is which vertices to merge at the coarsening stage of the algorithm. In this paper we introduce a new vertex similarity measure, as well as a coarsening scheme using it. The effectiveness is demonstrated by comparing the performance with other state-of-the-art hypergraph partitioners on problems from the Hypergraph Partitioning Archive
Available in PDF on arXiv.org.
Hypergraph Partitioning Archive
Fall 2017, regularly updated
An archive containing best known partitions for 490+ hypergraphs. Researchers are encouraged to submit their solutions.
Warping TensorFaces: Preprocessing Images for Multilinear Analysis of Facial Image Ensembles
An exploratory study of using preprocessing to improve the quality of multilinear analysis of facial images for facial recognition. Multilinear analysis is motivated by the fact that images are results of multiple factors like light conditions, pose and expression. However, we are not able to observe these latent variables. Multilinear analysis takes advantage of tensor decomposition methods to utilize those hidden variables. In this project we use Active Appearance models to warp the training images before using them to train the model. This project is based on the paper Multilinear image analysis for facial recognition by Vasilescu et al.
Report (PDF), Code on GitHub
Implemented an interpreter of a subset of Python 2.7 from a yacc-able version of the full Python 2.7 grammar. Includes AST generator, type system, functions, recursion, scope management and more.
C++, flex, and bison.
For CPSC 8270 “Translation of programming languages”.
To Agile, or not to Agile: A Comparison of Software Development Methodologies
A deep dive into the history and evolution of Agile software development methodologies. The paper discusses origins and current state of the Agile development methods, as well as provides a comparison of most commonly used practices.
In collaboration with Justin Sybrandt for CPSC 8750 “Software Architecture”.
Available in PDF on arXiv.org
Air Traffic and Economic Development
Assembled a dataset containing quarterly employment and air traffic data for years 1993-2015 and created a set of tools to edit it, retrieve data from it and add more data to it.
In collaboration with Ehsan Sadrfaridpour for CPSC 8480 “Network Science”.
Code and Data, Poster