The IRL-derived speed-preference weight from keystroke data correlates with Parkinson's motor severity and improves prediction beyond raw typing speed.
arXiv preprint arXiv:2507.06326
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Inverse Reinforcement Learning for Interpretable Keystroke Biomarkers in Parkinson's Disease
The IRL-derived speed-preference weight from keystroke data correlates with Parkinson's motor severity and improves prediction beyond raw typing speed.