一、任務基礎

泰坦尼克號沉沒是歷史上最著名的沉船事故之一。1912年4月15日，在她的處女航中，泰坦尼克號在與冰山相撞後沉沒，在2224名乘客和機組人員中造成1502人死亡。這場聳人聽聞的悲劇震驚了國際社會，併為船舶制定了更好的安全規定。造成海難失事的原因之一是乘客和機組人員沒有足夠的救生艇。儘管倖存下沉有一些運氣因素，但有些人比其他人更容易生存，例如婦女，兒童和上流社會。在這個案例中我們將運用機器學習來預測哪些乘客可以倖免於悲劇。

資料集連結：https://pan.baidu.com/s/1bVnIM5JVZjib1znZIDn10g 。提取碼：1htm 。

讀取titanic_train資料集

import pandas

# 讀取資料集
titanic = pandas.read_csv('titanic_train.csv')
titanic.head(10)

檢視資料集前10行

特徵名詞解釋

二、資料預處理

可以看到Age列有缺失值(NaN)。一般來說，資料發生缺失的話有兩種處理方法，一種填充缺失值，一種直接捨棄這個特徵。這裡一般來說Age對結果是有較大影響的，我們可以對缺失值進行填充，這裡可以填充平均值 。 

# Age 缺失值填充
titanic['Age'] = titanic['Age'].fillna(titanic['Age'].median())
print(titanic.describe())

填充後檢視資料集的描述

機器學習演算法一般來說解決不了對字元的分類。因為我們是要對Survived這列‘’0‘’和"1"進行分類嘛。所以我們就要把"Sex"這一列的資料進行處理，把它對映為數值型。那我們就把"male"和“female”進行處理，分別用0和1替代。

# print(titanic['Sex'].unique())

# Replace all the occurences of male with the number 0.
# 用數字0替換所有出現的男性。
titanic.loc[titanic["Sex"] == "male", "Sex"] = 0
titanic.loc[titanic["Sex"] == "female", "Sex"] = 1

同時，我們也把"Embarked"這一列資料進行同樣的處理

# print(titanic['Embarked'].unique())

# Embarked：上船港口，有三個取值，C/S/Q，是文字形式，不利於分析，
# 故可能需要對映到數值的值，而且有2個缺失值
titanic['Embarked'] = titanic['Embarked'].fillna('S') # 缺失值填充為這一列的眾數'S'
titanic.loc[titanic["Embarked"] == "S", "Embarked"] = 0
titanic.loc[titanic["Embarked"] == "C", "Embarked"] = 1
titanic.loc[titanic["Embarked"] == "Q", "Embarked"] = 2

三、分類任務

首先使用線性迴歸演算法來進行分類

# Import the linear regression(迴歸) class
# 注意不要導錯庫 
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import KFold

# The columns we'll use to predict the target
predictors = ['Pclass', 'Sex', 'Age', 'SibSp', 'Parch', 'Fare', 'Embarked']

# Initialize our algorithm class
alg = LinearRegression()

# Generate cross validation folds for the titanic dataset. It return the row indices
# corresponding(相應的) to train and test.
# 為Titanic資料集生成交叉驗證摺疊。它返回與訓練和測試相對應的行索引。
# We set random_state to ensure we get the same splits every time we run this.
# kf = KFold(titanic.shape[0], n_folds=3, random_state=1) 寫法錯誤已被棄用

# 樣本平均分成3份，3折交叉驗證
kf = KFold(n_splits=3,shuffle=False, random_state=1)

# 注意這裡不是kf.split(titanic.shape[0])，會報如下錯誤：
# Singleton array array(891) cannot be considered a valid collection.

predictions = []
# 交叉驗證 劃分訓練集 驗證集
for train, test in  kf.split(titanic):
    # The predictors we're using the train the algorithm. Note how we only take
    # the rows in the train folds
    # 注意我們只得到訓練集的rows
    train_predictors = titanic[predictors].iloc[train, :]
    # The target we're using to train the algorithm.
    train_target = titanic['Survived'].iloc[train]
    # Training the algorithm using the predictors and target
    alg.fit(train_predictors, train_target)
    # We can now make predictions on the test fold
    test_predictions = alg.predict(titanic[predictors].iloc[test, :])
    predictions.append(test_predictions)

檢視線性迴歸準確率

import numpy as np

# The predictions are in three separate numpy arrays. Concatenate them into one.
# We concatenate them on asix 0, as they only have one axis.
predictions = np.concatenate(predictions,axis=0)

# Map predictions to outcomes (only possible outcomes are 1 and 0)
predictions[predictions > .5] = 1  # 對映成分類結果 計算準確率
predictions[predictions <= .5] = 0

# 注意這一行與原始碼有出入
accuracy = sum(predictions==titanic['Survived'])/len(predictions)

# 驗證集的準確率 
print(accuracy)

得到的準確率為

0.7833894500561167

對於一個二分類問題來說，這個準確率似乎不太行，接下來用邏輯迴歸演算法試下

from sklearn.model_selection import cross_val_score
from sklearn.linear_model import LogisticRegression

alg = LogisticRegression(random_state=1)
# Compute the accuracy score for all the cross validation folds,
# (much simper than what we did before!)
scores = cross_val_score(alg, titanic[predictors], titanic["Survived"], cv=3)
# Take the mean of the scores (because we have one for each fold)
print(scores.mean())

得到的準確率為，可以發現效果要好了一點。

0.8047138047138048

上面得到的結果都是對交叉驗證後的驗證集來進行分類，在實際結果中，應該使用測試資料集來進行分類。

讀取測試資料集並填充資料集，然後進行數值對映，與上面類似。

titanic_test = pandas.read_csv("test.csv")
titanic_test["Age"] = titanic_test["Age"].fillna(titanic["Age"].median())
titanic_test["Fare"] = titanic_test["Fare"].fillna(titanic_test["Fare"].median())
titanic_test.loc[titanic_test["Sex"] == "male", "Sex"] = 0 
titanic_test.loc[titanic_test["Sex"] == "female", "Sex"] = 1
titanic_test["Embarked"] = titanic_test["Embarked"].fillna("S")

titanic_test.loc[titanic_test["Embarked"] == "S", "Embarked"] = 0
titanic_test.loc[titanic_test["Embarked"] == "C", "Embarked"] = 1
titanic_test.loc[titanic_test["Embarked"] == "Q", "Embarked"] = 2

通過上面發現，似乎線性迴歸，邏輯迴歸這類演算法似乎不太行，那這次再用隨機森林演算法來試下(一般來說隨機森林演算法比線性迴歸和邏輯迴歸演算法的效果好一點)，注意隨機森林引數的變化。

from sklearn.model_selection import cross_val_score
from sklearn.model_selection import KFold
from sklearn.ensemble import RandomForestClassifier

predictors = ["Pclass", "Sex", "Age", "SibSp", "Parch", "Fare", "Embarked"]

# Initialize our algorithm with the default paramters
# n_estimators is the number of trees we want to make
# min_samples_split is the minimum number of rows we need to make a split
# min_samples_leaf is the minimum number of samples we can have at the place where a
# tree branch(分支) ends (the bottom points of the tree)
alg = RandomForestClassifier(random_state=1,
                             n_estimators=10,
                             min_samples_split=2,
                             min_samples_leaf=1)
# Compute the accuracy score for all the cross validation folds.  (much simpler than what we did before!)
kf = KFold(n_splits=3, shuffle=False, random_state=1)
scores = cross_val_score(alg, titanic[predictors], titanic["Survived"], cv=kf)

# Take the mean of the scores (because we have one for each fold)
print(scores.mean())

準確率為：

0.7856341189674523

發現準確率還是不太行。在機器學習中，調整引數也是非常重要的，一般通過引數的調整來對模型進行優化。這次調整隨機森林的引數。

alg = RandomForestClassifier(random_state=1,
                             n_estimators=100,
                             min_samples_split=4,
                             min_samples_leaf=2)
# Compute the accuracy score for all the cross validation folds.  (much simpler than what we did before!)
kf = KFold(n_splits=3, shuffle=False, random_state=1)
scores = cross_val_score(alg, titanic[predictors], titanic["Survived"], cv=kf)

# Take the mean of the scores (because we have one for each fold)
print(scores.mean())

得到的準確率為：

0.8148148148148148

未完待續。