Hyunseok (Peter) Jang
University of Toronto
Product Engineer (Co-op) @ Survalent
hyunseok [dot] jang [at] mail [dot] utoronto [dot] ca
I am an undergraduate student at the University of Toronto studying Mathematics, Statistics, and Physics. Currently, I am at Survalent working on predictive models for electric power distribution systems (ADMS and SCADA).
Previously, I worked on national research projects in South Korea funded by the Ministry of Trade, Industry and Energy.
My research interests lie in Bayesian statistics and machine learning. I'm no master (yet), but with my limited toolbox, I've been working on probabilistic forecasting, model interpretations (XAI), generative models, and graph-based models. I hope to elevate my skills to the point where I can propose new learning algorithms and ML methods.
My current research questions are:
- How can we train and forecast time-series data better?
- How can we interpret forecast results better?
- Can we make stronger forecasts using pre-trained/generative models, perhaps develop a DALL-E or GPT-3 for time-series data?