Profile Picture
kaan@gatech.edu CV

Hi I'm Kaan Sancak

I am a CS Ph.D. student in the Collage of Computing at Georgia Tech advised by Umit V. Catalyurek.

My research interest include High Performance Computing, Graph Machine Learning and Deep Learning. My recent works focuses on scalability of Graph Neural Networks(GNNs), and interpretability of GNNs. My goal is to improve the performance and the efficiency of GNN training and make them more interpretable.

News

Education

Pic
Aug 2019 - Present

Georgia Institute of Technology

Ph.D. Computer Science
Advisor: Umit V. Catalyurek
Pic
Aug 2015 - May 2019

Bilkent University

B.S. Computer Science
Summa Cum Laude (Ranked 4 among 231)

Publications

MG-GCN: Scalable Multi-GPU GCN Training Framework
Kaan Sancak*, M. Fatih Balin*, , Umit V. Catalyurek
To be appear in ICPP22
test Abstract
test Paper
test BibTex
test Github
ElGA: Elastic and Scalable Dynamic Graph Analysis
Kasimir Gabert, Kaan Sancak, M. Yusuf Ozkaya, Ali Pinar, Umit V. Catalyurek
SC21: International Conference for High Performance Computing, Networking, Storage and Analysis
test Abstract
test BibTex
On Symmetric Rectilinear Matrix Partitioning
Abdurrahman Yasar, M. Fatih Balin, Xiaojjing An, Kaan Sancak, Umit V. Catalyurek
ACM Journal of Experimental Algorithmics (JEA)
test Abstract
test Paper
test BibTex
test Github
test Webpage

Selected Projects

SARMA

SARMA (SpatiAl Rectilinear Matrix pArtitioning) is a template-based, header only, library for spatial rectilinear partitioning. The main goal of this library to introduce novel symmetric rectilinear partitioning algorithms.

test GitHub

PathwayMapper

PathwayMapper is a web based pathway curation tool for interactive creation, editing, and sharing. Remote users to collaborate and concurrently modify pathways. My contribution consists of implementing collaborative editing features using ShareDB with built-in conflict resolution.

test GitHub

Newt

Newt is a sample application to visualize and edit the pathway models represented by process description and activity flow languages of SBGN or in simple interaction format. My contribution consists of implementing different features and multiple bug fixes.

test GitHub

cBioPortal

The cBioPortal for Cancer Genomics provides visualization, analysis, and download of large-scale cancer genomics datasets.

test GitHub

WanderTime

A full travel itinerary generator application that creates trips considering both transportation and accommodation.

test Demo
test Best demonstration award among 30 projects.

Service

Reviewer

2021

IEEE Transactions on Big Data | TBD

Other Experiences

Awards and Certificates

2019

Summa Cum Laude

Awarded by Department of Computer Science, Bilkent University
2016-2019

High Honor Student

Awarded 7 times for achieving a GPA above 3.50/4.00, Bilkent University
2017-2018

IBM University Ambassador

Awarded by IBM
2018

Deep Learning for Computer Vision Completion Certificate

Awarded by NVDIA DLI.
2015

Honor Student

Awarded by Department of Computer Science, Bilkent University Turkey.

Skills

Programming Languages

C++

C++

Java

Java

Python

Python

Frameworks & Libraries

PyTorch

PyTorch

Tensorflow

Tensorflow

DGL

DGL

CUDA

CUDA

Spark & Graphx

Spark & Graphx

Flink

Flink

Web Development

Javascript

Javascript

HTML

HTML

CSS

CSS

React

React

Svelte

Svelte

Other

Microsoft Office

Microsoft Office

LaTeX

LaTeX