Overview
Grace Wahba is a statistician (University of Wisconsin-Madison, emerita), a foundational figure in spline smoothing, reproducing-kernel Hilbert space (RKHS) methods, and the statistical theory connecting regularization to Bayesian estimation.
Key Contributions / Features
- Spline smoothing as Bayesian estimation (Wahba 1978): the equivalence between (generalized) spline smoothing and Bayes estimation under a partially improper prior.
- Generalized cross-validation (GCV) (Craven-Wahba 1979) for choosing the smoothing parameter, and the RKHS/representer-theorem framework (Kimeldorf-Wahba 1970, 1971).
- Spline Models for Observational Data (1990); foundational work bridging smoothing splines, kriging, and support-vector/kernel methods.
Related