|
Pavlos Constas-Malevanets I'm a first-year undergraduate student in Engineering Science at the University of Toronto, passionate about research and engineering innovation. My projects range from exploring ultra-wideband and single-photon imaging techniques to implementing a reinforcement learning method focused on medication adjustment to minimize speech disfluency. I've also gained industry experience at Advanced Mineral Technological Laboratory, where I have applied computer vision, robotics, chemical analysis, and repaired a long-standing mass spectrometer system failure. |
|
|
Toward a Reinforcement-Learning-Based System for Adjusting Medication to Minimize Speech Disfluency
Pavlos Constas, Vikram Rawal, Matthew Honorio Oliveira, Andreas Constas, Aditya Khan, Kaison Cheung, Najma Sultani, Carrie Chen, Micol Altomare, Michael Akzam, Jiacheng Chen, Vhea He, Lauren Altomare, Heraa Murqi, Asad Khan, Nimit Amikumar Bhanshali, Youssef Rachad, Michael Guerzhoy arXiv, 2024 arXiv A reinforcement learning algorithm is, under some circumstances, able to converge to a good medication regime that may help the simulated patient with their mental health-related speech disfluency, and adjust the medication and the dosages in response to zero-cost frequent measurement of the fluency of the patient. |
|
Rowelometer
Pavlos Constas-Malevanets Project, 2023 I propose a robust and inexpensive device for measuring the speed of a rowing shell and the stroke rate of the rowers. I used a GPS receiver, an accelerometer, and a microcontroller to measure the speed of the shell and the stroke rate of the rowers. The method behind the technology relies on Kalman Filtering to combine the GPS and accelerometer data to produce a more accurate measurement of the shell's speed. The method is accurate and has been thoroughly tested by comparing it alongside the industry standard. I used Altium Designer to design the PCB and the microcontroller code is written in C. The total cost of the device is less than $100. |