ML(0816_day9) - 결정 트리

결정 트리¶

In [1]:

import numpy as np
import matplotlib.pyplot as plt

from sklearn.datasets import load_iris
from sklearn.tree import DecisionTreeClassifier

In [93]:

iris = load_iris()

X = iris.data[:, 2:] # petal length, petal width
y = iris.target

In [94]:

tree_clf = DecisionTreeClassifier(max_depth=2, random_state=42)
tree_clf.fit(X, y)

Out[94]:

DecisionTreeClassifier(max_depth=2, random_state=42)

In [95]:

from sklearn.tree import export_graphviz

In [96]:

export_graphviz(tree_clf, 
                out_file="iris_tree.dot", 
                class_names=iris.target_names, 
                feature_names=iris.feature_names[2:], 
                rounded=True, 
                filled=True)

In [97]:

import graphviz

with open("iris_tree.dot") as f:
    dot_graph = f.read()
graphviz.Source(dot_graph)

Out[97]:

In [98]:

tree_clf.predict_proba([[5, 1.5]])

Out[98]:

array([[0.        , 0.90740741, 0.09259259]])

In [99]:

tree_clf.predict([[5, 1.5]])

Out[99]:

array([1])

In [100]:

X.shape

Out[100]:

(150, 2)

• 특성 중요도

In [104]:

iris.feature_names[2:]

Out[104]:

['petal length (cm)', 'petal width (cm)']

In [101]:

tree_clf.feature_importances_

Out[101]:

array([0.56199095, 0.43800905])

• 회귀 트리

In [84]:

np.random.seed(42)

In [85]:

m = 200
X = np.random.rand(m, 1)
y = 4 * (X-0.5) ** 2
y = y + np.random.randn(m, 1) / 10

In [86]:

from sklearn.tree import DecisionTreeRegressor

In [87]:

tree_reg = DecisionTreeRegressor(max_depth=2, random_state=42)
tree_reg.fit(X, y)

Out[87]:

DecisionTreeRegressor(max_depth=2, random_state=42)

In [88]:

export_graphviz(tree_reg,
                out_file = "regression_tree.dot",
               feature_names=["x1"],
               rounded=True,
               filled=True)

In [89]:

with open("regression_tree.dot") as f:
    dot_graph = f.read()
graphviz.Source(dot_graph)

Out[89]:

In [90]:

tree_reg.predict([[0.6]])

Out[90]:

array([0.11063973])

In [ ]: