tf神经网络模型预测泰坦尼克号生还

前言：

数据集找我

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import tensorflow as tf
train_data = pd.read_csv("./datasets/train.csv")
train_data.head(10)
test_data = pd.read_csv("./datasets/test.csv")
test_data.head(10)
Y_train = train_data["Survived"]
features = ["Pclass", "Sex", "SibSp", "Parch"]
X_train= pd.get_dummies(train_data[features])
X_test = pd.get_dummies(test_data[features])

from sklearn.model_selection import train_test_split
X_train, X_val, Y_train, Y_val = train_test_split(X_train, Y_train, test_size=0.3,random_state=100)
#定义神经网络模型
print(X_train.head(10));
print(Y_train.head(10));
model = tf.keras.Sequential( name = 'Titanic')  #定义模型
model.add(tf.keras.layers.InputLayer(input_shape = (5)))  #定义输入层
#model.add(tf.keras.layers.Flatten())
model.add(tf.keras.layers.Dense(4, activation='relu'))  #定义隐藏层
model.add(tf.keras.layers.Dense(1, activation='sigmoid'))  #定义输出层
model.summary()  #显示模型摘要
#编译模型
model.compile(optimizer = 'adam',
             loss = 'binary_crossentropy',
             metrics = ['accuracy'])
#训练模型
epochs = 30
print(X_train,Y_train);
h = model.fit(X_train,Y_train,
          batch_size = 4,
          validation_data = (X_val,Y_val),
          epochs =epochs)
# # 绘制损失函数曲线
plt.figure(figsize=(8,5))
plt.plot(h.history['loss'], label = 'loss')
plt.plot(h.history['val_loss'], label = 'val_loss')
plt.xticks(range(epochs),range(1,epochs+1))
plt.xlabel('epoch')
plt.ylabel('loss')
plt.legend()
plt.show()
# 绘制准确率曲线
plt.figure(figsize=(8,5))
plt.plot(h.history['accuracy'], label = 'acc')
plt.plot(h.history['val_accuracy'], label = 'val_acc')
plt.xticks(range(epochs),range(1,epochs+1))
plt.xlabel('epoch')
plt.ylabel('accuracy')
plt.legend()
plt.show()
#
predictions = model.predict(X_test)
output = pd.DataFrame({
    
    'PassengerId': test_data.PassengerId, 'Survived': predictions.ravel()})
output.to_csv('my_submission.csv', index=False)