分析手法

複数の分類器で一気に比較

2018年1月19日 by 河副太智 Leave a Comment

複数の分類器

import pandas as pd
from sklearn.model_selection import train_test_split

df = pd.read_csv('train.csv')

df = df.drop(['Cabin','Name','PassengerId','Ticket'],axis=1)
train_X = df.drop('Survived', axis=1)
train_y = df.Survived
(train_X, test_X ,train_y, test_y) = train_test_split(train_X, train_y, test_size = 0.3, random_state = 666)


#決定木
from sklearn.tree import DecisionTreeClassifier
ki = DecisionTreeClassifier(random_state=0).fit(train_X, train_y)
print(ki.score(train_X,train_y))



#ランダムフォレスト
from sklearn.ensemble import RandomForestClassifier
mori = RandomForestClassifier(random_state=0).fit(train_X,train_y)
print(mori.score(train_X,train_y))


#ロジスティック回帰
from sklearn.linear_model import LogisticRegression
logi = LogisticRegression(C=100).fit(train_X,train_y)
print(logi.score(train_X,train_y))


#KNN
from sklearn.neighbors import KNeighborsClassifier
KNN = KNeighborsClassifier(4).fit(train_X,train_y)
print(KNN.score(train_X,train_y))

#SVC
from sklearn.svm import SVC
svc = SVC(probability=True).fit(train_X,train_y)
print(svc.score(train_X,train_y))

#AdaBoostClassifier
from sklearn.ensemble import AdaBoostClassifier
ada = AdaBoostClassifier().fit(train_X,train_y)
print(ada.score(train_X,train_y))

#GradientBoostingClassifier
from sklearn.ensemble import GradientBoostingClassifier
gra = GradientBoostingClassifier().fit(train_X,train_y)
print(gra.score(train_X,train_y))

#GaussianNB
from sklearn.naive_bayes import GaussianNB
gaus = GaussianNB().fit(train_X,train_y)
print(gaus.score(train_X,train_y))

#LinearDiscriminantAnalysis
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
lda = LinearDiscriminantAnalysis().fit(train_X,train_y)
print(lda.score(train_X,train_y))

#QuadraticDiscriminantAnalysis
from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis
qua = QuadraticDiscriminantAnalysis().fit(train_X,train_y)
print(qua.score(train_X,train_y))

import pandas as pd

from sklearn.model_selection import train_test_split

df = pd.read_csv('train.csv')

df = df.drop(['Cabin','Name','PassengerId','Ticket'],axis=1)

train_X = df.drop('Survived', axis=1)

train_y = df.Survived

(train_X, test_X ,train_y, test_y) = train_test_split(train_X, train_y, test_size = 0.3, random_state = 666)

#決定木

from sklearn.tree import DecisionTreeClassifier

ki = DecisionTreeClassifier(random_state=0).fit(train_X, train_y)

print(ki.score(train_X,train_y))

#ランダムフォレスト

from sklearn.ensemble import RandomForestClassifier

mori = RandomForestClassifier(random_state=0).fit(train_X,train_y)

print(mori.score(train_X,train_y))

#ロジスティック回帰

from sklearn.linear_model import LogisticRegression

logi = LogisticRegression(C=100).fit(train_X,train_y)

print(logi.score(train_X,train_y))

#KNN

from sklearn.neighbors import KNeighborsClassifier

KNN = KNeighborsClassifier(4).fit(train_X,train_y)

print(KNN.score(train_X,train_y))

#SVC

from sklearn.svm import SVC

svc = SVC(probability=True).fit(train_X,train_y)

print(svc.score(train_X,train_y))

#AdaBoostClassifier

from sklearn.ensemble import AdaBoostClassifier

ada = AdaBoostClassifier().fit(train_X,train_y)

print(ada.score(train_X,train_y))

#GradientBoostingClassifier

from sklearn.ensemble import GradientBoostingClassifier

gra = GradientBoostingClassifier().fit(train_X,train_y)

print(gra.score(train_X,train_y))

#GaussianNB

from sklearn.naive_bayes import GaussianNB

gaus = GaussianNB().fit(train_X,train_y)

print(gaus.score(train_X,train_y))

#LinearDiscriminantAnalysis

from sklearn.discriminant_analysis import LinearDiscriminantAnalysis

lda = LinearDiscriminantAnalysis().fit(train_X,train_y)

print(lda.score(train_X,train_y))

#QuadraticDiscriminantAnalysis

from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis

qua = QuadraticDiscriminantAnalysis().fit(train_X,train_y)

print(qua.score(train_X,train_y))

Out[]:
0.982343499197
0.967897271268
0.807383627608
0.796147672552
0.886035313002
0.837881219904
0.898876404494
0.796147672552
0.799357945425
0.813804173355

決定木

2018年1月16日 by 河副太智 Leave a Comment

決定木

train_X = df.drop('Survived', axis=1)
train_y = df.Survived
(train_X, test_X ,train_y, test_y) = train_test_split(train_X, train_y, test_size = 0.3, random_state = 666)

#決定木
clf = DecisionTreeClassifier(random_state=0)
clf = clf.fit(train_X, train_y)
pred = clf.predict(test_X)

#決定木のモデルスコア
fpr, tpr, thresholds = roc_curve(test_y, pred, pos_label=1)
auc(fpr, tpr)
accuracy_score(pred, test_y)

train_X = df.drop('Survived', axis=1)

train_y = df.Survived

(train_X, test_X ,train_y, test_y) = train_test_split(train_X, train_y, test_size = 0.3, random_state = 666)

#決定木

clf = DecisionTreeClassifier(random_state=0)

clf = clf.fit(train_X, train_y)

pred = clf.predict(test_X)

#決定木のモデルスコア

fpr, tpr, thresholds = roc_curve(test_y, pred, pos_label=1)

auc(fpr, tpr)

accuracy_score(pred, test_y)

ランダムフォレスト

2018年1月16日 by 河副太智 Leave a Comment

ランダムフォレスト

train_X = df.drop('Survived', axis=1) train_y = df.Survived (
train_X, test_X ,train_y, test_y
) = train_test_split(train_X, train_y, test_size = 0.3, random_state = 666)


#ランダムフォレスト
from sklearn.ensemble import RandomForestClassifier
clf = RandomForestClassifier(random_state=0)
clf = clf.fit(train_X, train_y)
pred = clf.predict(test_X)
fpr, tpr, thresholds = roc_curve(test_y, pred, pos_label=1)
auc(fpr, tpr)
accuracy_score(pred, test_y)

train_X = df.drop('Survived', axis=1) train_y = df.Survived (

train_X, test_X ,train_y, test_y

) = train_test_split(train_X, train_y, test_size = 0.3, random_state = 666)

#ランダムフォレスト

from sklearn.ensemble import RandomForestClassifier

clf = RandomForestClassifier(random_state=0)

clf = clf.fit(train_X, train_y)

pred = clf.predict(test_X)

fpr, tpr, thresholds = roc_curve(test_y, pred, pos_label=1)

auc(fpr, tpr)

accuracy_score(pred, test_y)

criterion : 分割基準。gini or entropyを選択。(デフォルトでジニ係数)
max_depth : 木の深さ。木が深くなるほど過学習し易いので、適当なしきい値を設定してあげる。
max_features：最適な分割をする際の特徴量の数
min_samples_split：分岐する際のサンプル数
random_state：ランダムseedの設定。seedを設定しないと、毎回モデル結果が変わるので注意。

公式ドキュメント
sklearn.tree.DecisionTreeClassifier — scikit-learn 0.19.1 documentation

どの要素が強く影響を与えているかを確認

import matplotlib.pyplot as plt
import numpy as np
%matplotlib inline

features = train_X.columns
importances = clf.feature_importances_
indices = np.argsort(importances)

plt.figure(figsize=(6,6))
plt.barh(range(len(indices)), importances[indices], color='b', align='center')
plt.yticks(range(len(indices)), features[indices])
plt.show()

import matplotlib.pyplot as plt

import numpy as np

%matplotlib inline

features = train_X.columns

importances = clf.feature_importances_

indices = np.argsort(importances)

plt.figure(figsize=(6,6))

plt.barh(range(len(indices)), importances[indices], color='b', align='center')

plt.yticks(range(len(indices)), features[indices])

plt.show()