Polynomial spline regression: Theory and Application
Spline (mechanical)
B-spline
DOI:
10.48550/arxiv.2212.14777
Publication Date:
2022-01-01
AUTHORS (2)
ABSTRACT
To deal with non-linear relations between the predictors and response, we can use transformations to make data look linear or approximately linear. In practice, however, transformation methods may be ineffective, it more efficient flexible regression techniques that automatically handle nonlinear behavior. One such method is Polynomial Spline (PS) regression. Because number of possible spline models many, strategies for choosing best one are required. This study investigates different (Polynomial based on Truncated Power, B-spline, P-Spline) in theoretical practical ways. We focus fundamental concepts as theoretically rich. particular, prediction using cross-validation (CV) rather than interpretation, polynomial splines challenging interpret. compare PS a real set conclude P-spline model best.
SUPPLEMENTAL MATERIAL
Coming soon ....
REFERENCES ()
CITATIONS ()
EXTERNAL LINKS
PlumX Metrics
RECOMMENDATIONS
FAIR ASSESSMENT
Coming soon ....
JUPYTER LAB
Coming soon ....