A Genetic Programming-Driven Data Fitting Method

杨利霞

A Genetic Programming-Driven

Data Fitting Method

5 M8 B/ E! W3 \5 n. A4 F


3 G" g- z2 H$ a# {. T/ fData fifitting is the process of constructing a curve, or a set of mathematical functions, that has
the best fifit to a series of data points. Different with constructing a fifitting model from same type of function,
: F  Z) i& b) m- ^4 ksuch as the polynomial model, we notice that a hybrid fifitting model with multiple types of function may have
1 S& B* G& Q9 Z" O# L! v/ k) Oa better fifitting result. Moreover, this also shows better interpretability. However, a perfect smooth hybrid
# e. D  }  r. Gfifitting model depends on a reasonable combination of multiple functions and a set of effective parameters.
That is a high-dimensional multi-objective optimization problem. This paper proposes a novel data fifitting
& W7 @. D8 G. s6 R( Imodel construction approach. In this approach, the model is expressed by an improved tree coding expression
) E8 U% P/ G7 B( v! \# M* T! xand constructed through an evolution search process driven by the genetic programming. In order to verify
' n5 J2 _6 ^6 ~the validity of generated hybrid fifitting model, 6 prediction problems are chosen for experiment studies. The
experimental results show that the proposed method is superior to 7 typical methods in terms of the prediction
accuracy and interpretability.
7 P: |- i% O* D: U+ p


& x: R! i1 H: h5 k1 U9 I/ M
' j0 W! s& n1 J8 O% ^* ?* G# U