Alexander Peplowski

Graduate Student at Mila - Quebec AI Institute

Hi, you found my website! Here you’ll find a showcase of my recent work.

Portfolio

GANs Without Convolutions
GANs Without Convolutions
Co-Author Jan 2021 - Apr 2021

In-depth reproduction, study, and extension of TransGAN, a GAN framework which eliminates the use of convolutional operations.

Visit
Inference of Public Transit Timeliness
Inference of Public Transit Timeliness
Co-Author May 2020 - Dec 2020

Probabilistic inference of the timeliness of bus routes using two-dimensional markov chains

Visit
AI Computer Vision for Mobile Robotics
AI Computer Vision for Mobile Robotics
Developer May 2020 - Dec 2020

Deep learning object detector models designed for a low-data, embedded environment

Details
Solar Radiation Forecasting from Satellite Imagery
Solar Radiation Forecasting from Satellite Imagery
Co-Author Jan 2020 - Apr 2020

3D CNN and RNN deep learning models for forecasting solar radiation

Visit
Low-Resource Machine Translation
Low-Resource Machine Translation
Co-Author Jan 2020 - Apr 2020

BERT masked language model embeddings, transformer encoder-decoder model, and back-translation for low-resource English-French translation.

Visit
CMC Electronics Augmented Global Positionning Receiver
CMC Electronics Augmented Global Positionning Receiver
Developer Dec 2014 - Sept 2019

A high-performance civil aviation precision approach navigation GPS receiver

Details
Pharmahacks Weekend Hackathon
Pharmahacks Weekend Hackathon
Co-Author Nov 2019

Identification and forecasting of pharmaceutical supply chain issues

Details

Experience

1
Research Intern
Hydro-Québec’s Research Institute (IREQ)

May 2020 - Dec 2020, Varennes, Quebec, Canada

IREQ is home to nearly 500 researchers, engineers and technicians who devote themselves to making Hydro Quebec’s energy system more efficient and intelligent.

Responsibilities:
  • Refined and deployed single-shot convolutional neural network object detectors on field robots
  • Improved training robustness in a low-data setting with dataset augmentation and saliency mapping

Teaching Assistant
University of Montreal

Sept 2020 - Dec 2020, Montreal, Quebec, Canada

Mila is based at the University of Montreal. The Mila course IFT6758 (Data Science) introduces students to concepts in data science.

Responsibilities:
  • Taught lab sections covering material relating to OpenCV, scikit-learn, hyperparameter searching, feature selection, Kaggle competitions and API
  • Evaluated student-submitted homework, exams and projects
2

3
Software Developer
CMC Electronics

Dec 2014 - Sept 2019, Montreal, Quebec, Canada

CMC Electronics designs and produces leading technology electronics products for the aviation and global positioning markets.

Responsibilities:
  • Developed and released production C++ and Ada code for real-time, safety-critical aircraft navigation systems
  • Optimized embedded software libraries to improve performance
  • Developed and debugged multi-threaded applications

Education

Sept 2019 - May 2021
M.Sc. in Machine Learning
Courses
  • Machine Learning
  • Data Science
  • Probabilistic Graphical Models
  • Algorithms
2013 - 2014
M.Eng. in Electrical Engineering, Control Systems
Courses
  • Control Systems
  • Mobile Robotics
  • Project Management
2009 - 2013
B. Eng. in Electrical Engineering (Honours)
Courses
  • Computer Vision
  • Operating Systems
  • Computer Engineering
  • Signals and Systems