Research
What shapes a scientific conclusion?
I study how evidence, assumptions, and computational choices influence what we can conclude from data — and how those conclusions can be evaluated and interpreted.
Four connected research questionsSelect a theme to explore
Shared concernWhat does a result depend on?
01 / EVIDENCE & ASSUMPTIONS
Different accounts, different results?
If the evidence supports more than one scientific account, which conclusions depend on the account used to train a model?
I compare alternative learning conditions and examine their effects on specific scientific quantities, including what is missed when full retraining is approximated.
Research in practice: controlled representation-learning experiments; AMR-genomics feasibility studies.