《Tensorflow基础泰坦尼克获救预测》-- 网易云课堂

 1 # -*- coding: utf-8 -*-
 2 
 3 import tensorflow as tf
 4 import pandas as pd
 5 import numpy as np
 6 
 7 data = pd.read_csv(‘train.csv‘)
 8 data = data[[‘Survived‘, ‘Pclass‘, ‘Sex‘, ‘Age‘, ‘SibSp‘, ‘Parch‘, ‘Fare‘, ‘Cabin‘, ‘Embarked‘]]
 9 
10 data[‘Age‘] = data[‘Age‘].fillna(data[‘Age‘].mean())
11 data[‘Cabin‘] = pd.factorize(data[‘Cabin‘])[0]
12 data.fillna(0, inplace=True)
13 data[‘Sex‘] = [1 if x==‘male‘ else 0 for x in data[‘Sex‘]]
14 data[‘p1‘] = np.array(data[‘Pclass‘]==1).astype(np.int32)
15 data[‘p2‘] = np.array(data[‘Pclass‘]==2).astype(np.int32)
16 data[‘p3‘] = np.array(data[‘Pclass‘]==3).astype(np.int32)
17 del data[‘Pclass‘]
18 data[‘e1‘] = np.array(data[‘Embarked‘]==‘S‘).astype(np.int32)
19 data[‘e2‘] = np.array(data[‘Embarked‘]==‘C‘).astype(np.int32)
20 data[‘e3‘] = np.array(data[‘Embarked‘]==‘Q‘).astype(np.int32)
21 del data[‘Embarked‘]
22 
23 data_train = data[[ ‘Sex‘, ‘Age‘, ‘SibSp‘, ‘Parch‘, ‘Fare‘, ‘Cabin‘, ‘p1‘, ‘p2‘, ‘p3‘, ‘e1‘, ‘e2‘, ‘e3‘]]
24 data_target = data[‘Survived‘].values.reshape(len(data), 1)
25 
26 x = tf.placeholder("float", shape=[None, 12])
27 y = tf.placeholder("float", shape=[None, 1])
28 
29 weight = tf.Variable(tf.random_normal([12, 1]))
30 bias = tf.Variable(tf.random_normal([1]))
31 output = tf.matmul(x, weight) + bias
32 pred = tf.cast(tf.sigmoid(output) > 0.5, tf.float32)
33 
34 loss = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(labels=y, logits=output))
35 train_step = tf.train.GradientDescentOptimizer(0.001).minimize(loss)
36 accurary = tf.reduce_mean(tf.cast(tf.equal(pred, y), tf.float32))
37 
38 data_test = pd.read_csv(‘test.csv‘)
39 data_test = data_test[[‘Pclass‘, ‘Sex‘, ‘Age‘, ‘SibSp‘, ‘Parch‘, ‘Fare‘, ‘Cabin‘, ‘Embarked‘]]
40 data_test[‘Age‘] = data_test[‘Age‘].fillna(data_test[‘Age‘].mean())
41 data_test[‘Cabin‘] = pd.factorize(data_test[‘Cabin‘])[0]
42 data_test.fillna(0, inplace=True)
43 data_test[‘Sex‘] = [1 if x==‘male‘ else 0 for x in data_test[‘Sex‘]]
44 data_test[‘p1‘] = np.array(data_test[‘Pclass‘]==1).astype(np.int32)
45 data_test[‘p2‘] = np.array(data_test[‘Pclass‘]==2).astype(np.int32)
46 data_test[‘p3‘] = np.array(data_test[‘Pclass‘]==3).astype(np.int32)
47 del data_test[‘Pclass‘]
48 data_test[‘e1‘] = np.array(data_test[‘Embarked‘]==‘S‘).astype(np.int32)
49 data_test[‘e2‘] = np.array(data_test[‘Embarked‘]==‘C‘).astype(np.int32)
50 data_test[‘e3‘] = np.array(data_test[‘Embarked‘]==‘Q‘).astype(np.int32)
51 del data_test[‘Embarked‘]
52 
53 test_label = pd.read_csv(‘gender_submission.csv‘)
54 test_label = np.reshape(test_label[‘Survived‘].values.astype(np.float32), (418,1))
55 
56 sess = tf.Session()
57 sess.run(tf.global_variables_initializer())
58 loss_train = []
59 train_acc = []
60 test_acc = []
61 
62 data_train = data_train.values
63 for i in range(25000):
64     index = np.random.permutation(len(data_target))
65     data_train = data_train[index]
66     data_target = data_target[index]
67     for n in range(len(data_target)//100 + 1):
68         batch_xs = data_train[n*100:n*100+100]
69         batch_ys = data_target[n*100:n*100+100]
70         sess.run(train_step, feed_dict={x:batch_xs, y:batch_ys})
71         
72     if i%1000 == 0:
73         loss_temp = sess.run(loss, feed_dict={x:batch_xs, y:batch_ys})
74         loss_train.append(loss_temp)
75         train_acc_temp = sess.run(accurary, feed_dict={x:batch_xs, y:batch_ys})
76         train_acc.append(train_acc_temp)
77         test_acc_temp = sess.run(accurary, feed_dict={x:data_test, y:test_label})
78         test_acc.append(test_acc_temp)
79         print(loss_temp,train_acc_temp,test_acc_temp)
80         
81 import matplotlib.pyplot as plt
82 
83 plt.plot(loss_train, ‘k-‘)
84 plt.title(‘train loss‘)
85 plt.show()
86 
87 plt.plot(train_acc, ‘b-‘, label=‘train_acc‘)
88 plt.plot(test_acc, ‘r--‘, label=‘test_acc‘)
89 plt.title(‘train and test accuracy‘)
90 plt.legend()
91 plt.show()