Kaggle挑戦記

KaggleのTitanicに参加した記録。 データを目で見て分かったのは、

• 兄弟の少ない子供は比較的助かった
• 兄弟の多い子供は比較的助からなかった

ということぐらいでしたが、機械学習で分類器を作った結果（データを読み込ませてSVMで分類させただけという何の工夫もないやり方ですが・・・）、0.77990というスコアになりました。そのうち細かいチューニングもして再チャレンジしてみます。

import pandas as pd
import matplotlib.pyplot as plt
%matplotlib inline

train = pd.read_csv("train.csv")
train.head()

	PassengerId	Survived	Pclass	Name	Sex	Age	SibSp	Parch	Ticket	Fare	Cabin	Embarked
0	1	0	3	Braund, Mr. Owen Harris	male	22.0	1	0	A/5 21171	7.2500	NaN	S
1	2	1	1	Cumings, Mrs. John Bradley (Florence Briggs Th...	female	38.0	1	0	PC 17599	71.2833	C85	C
2	3	1	3	Heikkinen, Miss. Laina	female	26.0	0	0	STON/O2. 3101282	7.9250	NaN	S
3	4	1	1	Futrelle, Mrs. Jacques Heath (Lily May Peel)	female	35.0	1	0	113803	53.1000	C123	S
4	5	0	3	Allen, Mr. William Henry	male	35.0	0	0	373450	8.0500	NaN	S

train = train.drop(["PassengerId","Name","Ticket","Cabin"],axis=1)
train.head()

	Survived	Pclass	Sex	Age	SibSp	Parch	Fare	Embarked
0	0	3	male	22.0	1	0	7.2500	S
1	1	1	female	38.0	1	0	71.2833	C
2	1	3	female	26.0	0	0	7.9250	S
3	1	1	female	35.0	1	0	53.1000	S
4	0	3	male	35.0	0	0	8.0500	S

# ここの内容は分類器には関係ない。データを理解するためにやっただけ。

import matplotlib.gridspec as gridspec

Survive = train[train["Survived"]==1]
Dead = train[train["Survived"]==0]

transparency = 0.1

plt.figure(figsize=(16, 5))
# 兄弟、配偶者の数
plt.subplot(1,2,1)
plt.scatter(Survive["Age"], Survive["SibSp"], c="blue", alpha=transparency, label="Survived")
plt.scatter(Dead["Age"], Dead["SibSp"], c="red", alpha=transparency, label="Dead")
plt.xlabel("Age")
plt.ylabel("SibSp")
plt.legend()

plt.subplot(1,2,2)
# 親、子供の数
plt.scatter(Survive["Age"], Survive["Parch"], c="blue", alpha=transparency, label="Survived")
plt.scatter(Dead["Age"], Dead["Parch"], c="red", alpha=transparency, label="Dead")
plt.xlabel("Age")
plt.ylabel("Parch")
plt.legend()
plt.show()

# データ欠損数を確認
train.isnull().sum()

Survived      0
Pclass        0
Sex           0
Age         177
SibSp         0
Parch         0
Fare          0
Embarked      2
dtype: int64

# 欠損値を埋める候補
print(train["Age"].mean())
# print(train["Embarked"].mode)
print(train["Embarked"].value_counts())

29.6991176471
S    644
C    168
Q     77
Name: Embarked, dtype: int64

# 欠損値を埋める
train["Age"] = train["Age"].fillna(train["Age"].mean())
train["Embarked"] = train["Embarked"].fillna("S")

# データ欠損数を確認
train.isnull().sum()

Survived    0
Pclass      0
Sex         0
Age         0
SibSp       0
Parch       0
Fare        0
Embarked    0
dtype: int64

# EmbarkedとAgeを数値に変換
train["Sex"] = train["Sex"].map({"male": 0, "female": 1})
train["Embarked"] = train["Embarked"].map({"C":0,"Q":1,"S":2})

# AgeとFareを正規化
train["Age"] = (train["Age"]-train["Age"].mean())/(train["Age"].max()-train["Age"].min())
train["Fare"] = (train["Fare"]-train["Fare"].mean())/(train["Fare"].max()-train["Fare"].min())

# testの読み込み
test = pd.read_csv("test.csv")
test["Sex"] = test["Sex"].map({"male": 0, "female": 1})
test["Embarked"] = test["Embarked"].map({"C":0,"Q":1,"S":2})
test["Age"] = test["Age"].fillna(test["Age"].mean())
test["Fare"] = test["Fare"].fillna(test["Fare"].mean())
test["Age"] = (test["Age"]-test["Age"].mean())/(test["Age"].max()-test["Age"].min())
test["Fare"] = (test["Fare"]-test["Fare"].mean())/(test["Fare"].max()-test["Fare"].min())
test = test.drop(["Name", "Ticket", "Cabin"], axis=1)

# Machine learning で使うデータとセットに分ける。
X_train = train.drop("Survived", axis=1)
Y_train = train["Survived"]
X_test  = test.drop("PassengerId", axis=1).copy()
X_train.shape, Y_train.shape, X_test.shape

((891, 7), (891,), (418, 7))

# 学習
from sklearn.svm import SVC
from sklearn.model_selection import cross_val_score
from sklearn.model_selection import KFold

kfold = KFold(n_splits=10, random_state=42)
result = cross_val_score(SVC(),X_train,Y_train, cv = kfold, scoring = "accuracy")
print(result)

[ 0.78888889  0.84269663  0.76404494  0.85393258  0.79775281  0.80898876
  0.78651685  0.7752809   0.85393258  0.79775281]

classifier = SVC()

classifier.fit(X_train, Y_train)
result = classifier.predict(X_test)

#  提出する
submission = pd.DataFrame({
        "PassengerId": test["PassengerId"],
        "Survived": result
    })

submission.to_csv("submission.csv", index=False)