My ikigai
My love to applied mathematics is rooted in broadly defined statistics and machine learning - inferring models of the world and their parameters, as well as rigorously quantifying uncertainty around them from data around us. My primary focus is on combining sequential decision making and causality.
Sequential decision making appeals to me for several reasons: (1) it reflects the dynamic nature of the real world, where decisions are made iteratively over time (2) it allows for improved efficiency relative to static scenarios by enabling adaptive changes as new information becomes available (3) it is about gathering useful information through interaction with environment - facilitating active sampling (4) gives formalism of long-term goals.
Causality fascinates me, due to its power to formalize knowledge and structure reasoning by: (1) distinguishing between observing naturally ocurring processes and interventions (2) providing a framework for understanding how changes in one variable can bring about changes in another (3) it offers tools to explore counterfactual queries, which allow us to reason precisely about alternative choices.
I strongly believe that the interplay between sequential decision making and causality represents a crucial step towards human-AI collaboration and AGI. Sequential decision making offers the capacity for adaptive, goal-oriented learning, while causality provides a rigorous foundation for understanding and reasoning about change. Together, they chart a promising path toward creating systems that are not only reactive but also proactive, reasoning agents capable of navigating the complexity of the real world.
My research obsessions include:
- CausalML
- CausalRL
- Bandits
- Causal Inference
- Active Learning
You can find my CV on the respective subpage.
My stack:
- Economics (Bachelor's level)
- Pure mathematics (Bachelor's level)
- Probability and statistics (Master's level)
- Machine learning (applied and theoretical)
- Causality (SCM & PO)
- Sequential decision making (bandits, MDP, active learning)
- Programming for data science purposes (Python, R, SQL, Git)
- Strong presentation, communication and storytelling skills
My name Drążkowski reads as "'ɣubɛrt ˈdrɔ̃ʐkɔfski" \approx in English "Hoo-bert Drahnzh-koff-ski", where "Drah" is like "draw", but with a slightly more open "ah" sound, and "nzh" is like the "s" in "treasure". I do not take responsibility for any tongue injuries.