标题: A Genetic Programming-Driven Data Fitting Method [打印本页] 作者: 杨利霞 时间: 2020-11-10 16:01 标题: A Genetic Programming-Driven Data Fitting Method
A Genetic Programming-Driven
* d% z9 e8 S% J% m
Data Fitting Method
/ v# `# o) M. F' o
+ L( h% T& w5 T( D4 e5 D6 _: i
; ^: Z3 p; K9 i- Y- M3 R! ~9 \% F9 W
Data fifitting is the process of constructing a curve, or a set of mathematical functions, that has( s w- b" Y5 z/ I4 x
the best fifit to a series of data points. Different with constructing a fifitting model from same type of function, 5 {2 ^! N. a0 [5 o, Osuch as the polynomial model, we notice that a hybrid fifitting model with multiple types of function may have( t. b. y9 l) n
a better fifitting result. Moreover, this also shows better interpretability. However, a perfect smooth hybrid 9 A/ K$ c9 }' ?fifitting model depends on a reasonable combination of multiple functions and a set of effective parameters.( O. O. [2 v6 u3 [& e4 Y8 s
That is a high-dimensional multi-objective optimization problem. This paper proposes a novel data fifitting 0 @( H4 h7 G# o6 ymodel construction approach. In this approach, the model is expressed by an improved tree coding expression' ?4 I% M% f1 T9 N/ K. z# N! ^
and constructed through an evolution search process driven by the genetic programming. In order to verify . W5 ?0 o1 u6 D: r% l4 u- Y* Ithe validity of generated hybrid fifitting model, 6 prediction problems are chosen for experiment studies. The * e$ U' z/ M3 o! t, U P7 uexperimental results show that the proposed method is superior to 7 typical methods in terms of the prediction6 u8 y: c9 n! y2 K$ G b8 U
accuracy and interpretability. / X2 y. @& D% F! |, @4 ]
* h2 Z. U# {* U" O
" c. X3 [2 e3 V8 |: i2 d3 }; r. T! o2 c/ e" R/ ], t a& f! \) f3 }; T
