xref: /petsc/doc/manualpages/MANSECHeaders/PetscRegressor (revision 6bfab51239a1d021a2781a42e04752bb50d6082e)
134b254c5SRichard Tran Mills# Regression Analysis and Classification (PetscRegressor)
234b254c5SRichard Tran Mills
334b254c5SRichard Tran MillsThe Regression Analysis and Classification (`PetscRegressor`) component provides a simple interface for supervised
4*789736e1SBarry Smithstatistical (or machine) learning regression (prediction of continuous numerical values, including least squares with `PETSCREGRESSORLINEAR`) or classification (prediction
534b254c5SRichard Tran Millsof discrete labels or categories) tasks.
634b254c5SRichard Tran Mills
734b254c5SRichard Tran Mills`PetscRegressor` internally employs `Tao` (or `KSP` for a few, specialized cases) to solve the underlying numerical
834b254c5SRichard Tran Millsoptimization problems. `PetscRegressor` users can set `Tao` options or otherwise directly manipulate the underlying
934b254c5SRichard Tran Mills`Tao` context, which can be accessed via `PetscRegressorGetTao()`. User guide chapter: [](ch_regressor).
10