A Genetic Programming-Driven Data Fitting Method

杨利霞

A Genetic Programming-Driven

Data Fitting Method

1 K+ Y; S. J' {+ `& h9 r$ NData 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,
such as the polynomial model, we notice that a hybrid fifitting model with multiple types of function may have
a better fifitting result. Moreover, this also shows better interpretability. However, a perfect smooth hybrid
4 a( p7 `8 P% y! f1 D* ]fifitting model depends on a reasonable combination of multiple functions and a set of effective parameters.
1 w; t5 x, u( x6 v! y9 K5 k# |That is a high-dimensional multi-objective optimization problem. This paper proposes a novel data fifitting
model construction approach. In this approach, the model is expressed by an improved tree coding expression
% t4 V' W/ |: ]( |. i: N; `* Vand constructed through an evolution search process driven by the genetic programming. In order to verify
; w/ @% P7 g& v! t0 q/ G# kthe validity of generated hybrid fifitting model, 6 prediction problems are chosen for experiment studies. The
) [$ O3 ~7 G) T9 b9 U  Kexperimental results show that the proposed method is superior to 7 typical methods in terms of the prediction
! ]1 S" `5 D4 }. s" ]6 |accuracy and interpretability.

1 R$ o1 r% z! k$ r8 s5 c

1 T( q3 W# J' @; d5 F8 z( n