ML Research Assistant

Researching uncertainty modeling and Transformers for biological risk assessment.

What I'm Doing

I am researching uncertainty modeling techniques and Transformer architectures for assessing invasive species risk. The work involves modeling non-stationary spatio-temporal sequences and estimating survival probabilities under varying environmental conditions.

Impact (Expected)

This research aims to improve how we quantify risk in biological invasions, potentially informing better decision-making for environmental protection.

What I'm Learning

I am gaining deep understanding of Transformer attention mechanisms, positional encodings, and probabilistic modeling. The research requires handling complex, heterogeneous data and formulating rigorous mathematical models for uncertainty quantification.

Key Highlights

• Researching species-conditioned temporal Transformer models for ballast-water–mediated marine bioinvasion risk in the North Atlantic, focusing on non-stationary spatio-temporal sequence modeling and probabilistic estimation of species survival risk.