Guys, hello! Today, the ProductStar team has prepared an article for you in which we examined the general principles of operation and areas of application of the decision tree.
Decision tree is a method for automatic analysis of large data sets. In this article, we will look at the general principles of operation and areas of application.
Decision trees are powerful data mining and predictive analytics tools. It helps with classification and regression problems.
, « …, ...». .
, , : , « 1000 , ».
( ), . , — .
, — , . :
— ;
— .
1950- . .
. ( ID3 4.5 5.0) , CART .
. — , — (node) (leaf). .
: , :
— , ;
— , .
, . , . , , .
. — , , — . , , , .
, . . , , .
, , .
?
, , . :
. . .
( ). .
. . , , .
— . ? , .
:
— ;
— , .
«» , - ( ), . , , . , , .
, , « ». S, :
n , Ci(i = 1..k);
m Aj(j = 1..m), .
:
S Ci, , . , , «» . , Ci. , .
S — . , . , .
S Ck. . Aj S, : a1, a2, …, ap), p — . S p (S1, S2, …, Sp), . , . , .
, . , .
: ID3, CART, C4.5, C5.0, NewId, ITrule, CHAID, CN2 . :
ID3 (Iterative Dichotomizer 3). . , ID3, . . .
C4.5. «» ID3, . 2008 Spring Science , C4.5 — Data Mining.
CART (Classification and Regression Tree). , . CART , .
4 :
.
.
.
.
.
, . , . — .
, - .
-
:
n — , Ni — i- , N — .
. , , . , .
Aj , .
. — . , :
Info(S) — , S , Info(Sa) — , , A.
Gain(A), . - « .
. , . — .
, . .
:
Q — , n — , pi — i- ( ).
0 1. 0, , . 1, , . , .
«» . , . - . . , .
. , . — .
, «».
:
. (, ). — . — . - .
. . .
. (, 7). .
, . , - . - , .
«» , . , 2-3 , .
— , , . — .
: NP- , , , . , 3 :
, .
: ( ) ( ).
, .
, , .
« » — . , .
, . , .
. , .
( ), .
:
. , « < 40 , ». .
, .
, «» , ( ).
.
.
, .
:
. , .
, - .
, - « », .
: , 100- .
, , .
?
. , .
:
. .
. ( ), (, ) ..
. .
. .
. .
. , - .
, . ProductStar vc -.
. , , .