Principal Differential Analysis
J. O. Ramsay · Encyclopedia of Statistical Sciences · 2004
Abstract Differential equation systems are often a natural way of describing sets of variables changing over time or space. For example, in controlling industrial systems, engineers consider the time course of a set of inputs and outputs for a system such as a chemical reactor with a view to developing useful feedback processes. Principal differential analysis uses functional data to estimate a system of linear differential equations that fit the data. The estimate minimizes an integrated squared error criterion. The coefficient functions that define the linear differential equation are estimated by expansions in terms of basis functions such as Fourier series orB‐splines, and the estimates are calculated rapidly and without iteration.
Cited by 1 works. doi:10.1002/0471667196.ess5023
Authors
References
- Functional Data Analysis
- Chaos
- Estimating Smooth Monotone Functions
- Applied Functional Data Analysis
- Functional Components of Variation in Handwriting
- Principal Differential Analysis: Data Reduction by Differential Operators
- Penalized regression with model‐based penalties
- Linear Ordinary Differential Equations
- Differential equation models for statistical functions