CV

ANTONIOS VALKANAS

Technical Skills: Pytorch, Python, Numpy, Pandas, Java, Git, Matlab, LaTeX

Knowledge in: Graph Learning, Deep Learning, Data Science, Machine Learning, Time Series

Languages: English (Fluent), French (Advanced), Greek (Fluent)

Master of Science - Electrical Engineering (Thesis) Sep 2019 to April 2021 (Expected) McGill University - GPA: 3.88

Research on deep learning for graph structures, multi-label classification and application to real world data.
Recipient of the prestigious Alexander Graham Bell Scholarship; a highly competitive Canada-wide scholarship awarded for academic excellence and research potential.

Bachelor of Engineering - Electrical (Honours) Sep 2015 to April 2019 McGill University - GPA: 3.79

Honours program included graduate classes on machine learning, multi-agent systems, computer vision.

Undergraduate Research Internship, McGill University May 2018 to Aug 2019

Developed a framework to apply computer vision techniques such as SR-CNN on electromagnetic problems for simulating conductor potentials.
Resulted in a peer-reviewed publication in IEEE conference Compumag 2019.
Started off as an awarded USRA project for summer research but later joined with undergrad thesis.

Academic Lecturer, McGill AI Society Sep 2019 to April 2020

Created the full content of a semester long extracurricular undergraduate level course to teach students modern statistical learning and deep learning techniques.
The course was recognized by the Dean of Engineering of McGill University and students who take it now have it show up as a zero credit course on their official transcript.

SDR Satellite Communication Project Jan 2019 to June 2019

Programmed a Software Defined Radio (SDR) receiver and built an antenna to receive weather satellite images from the NOAA satellite constellation.
Work published by an education journal elaborating on how this project can be used as a lab in introductory communication courses in engineering.

MIO-TCD Challenge Dec 2018

Open challenge consisted of two parts: classification and localization of vehicles from 786,702 road images.
Large amount of data (20 GB+) made training the models quite hard so Google cloud computing was used.
Architecture consisted of two separate CNNs; one for localization and one for classification.

McGill Greek Student Association April 2016 to Present

Vice President of the club, organized events where current students, alumni, faculty, the local greek community and the consul of Greece in Montreal.

Volunteer Tutor Sep 2014 to April 2015

Over 100 hours of tutoring math and science courses documented by Toronto District School Board (TDSB).