R1 RNN-心脏病预测

• 🍨 本文为🔗365天深度学习训练营 中的学习记录博客
• 🍖 原作者：K同学啊 | 接辅导、项目定制
• 🚀 文章来源：K同学的学习圈子

我的环境：

• 语言环境：python 3.8

• 编译器：jupyter notebook

• 深度学习环境：Pytorch

  torch == 2.1.0+cpu

  torchvision == 0.16.0+cpu

• 要求：

1. 本地读取并加载数据。
2. 了解RNN构建过程。
3. 测试集acc达到87%（89%更好）。

一、准备工作

import tensorflow as tf

import pandas as pd
import numpy as np

df = pd.read_csv("heart.csv")

df

	age	sex	cp	trestbps	chol	fbs	restecg	thalach	exang	oldpeak	slope	ca	thal	target
0	63	1	3	145	233	1	0	150	0	2.3	0	0	1	1
1	37	1	2	130	250	0	1	187	0	3.5	0	0	2	1
2	41	0	1	130	204	0	0	172	0	1.4	2	0	2	1
3	56	1	1	120	236	0	1	178	0	0.8	2	0	2	1
4	57	0	0	120	354	0	1	163	1	0.6	2	0	2	1
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
298	57	0	0	140	241	0	1	123	1	0.2	1	0	3	0
299	45	1	3	110	264	0	1	132	0	1.2	1	0	3	0
300	68	1	0	144	193	1	1	141	0	3.4	1	2	3	0
301	57	1	0	130	131	0	1	115	1	1.2	1	1	3	0
302	57	0	1	130	236	0	0	174	0	0.0	1	1	2	0

303 rows × 14 columns

#检查是否有空值
df.isnull().sum() 

age         0
sex         0
cp          0
trestbps    0
chol        0
fbs         0
restecg     0
thalach     0
exang       0
oldpeak     0
slope       0
ca          0
thal        0
target      0
dtype: int64

二、数据预处理

1. 划分训练集和测试集

from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import train_test_split

X = df.iloc[:,:-1]
y = df.iloc[:,-1]

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.1, random_state = 1)

X_train.shape, y_train.shape

((272, 13), (272,))

2. 标准化

#将每一列标准化为标准正态分布
sc = StandardScaler()
X_train = sc.fit_transform(X_train)
X_test = sc.transform(X_test)

X_train = X_train.reshape(X_train.shape[0], X_train.shape[1], 1)
X_test = X_test.reshape(X_test.shape[0], X_test.shape[1], 1)

X_train.shape, X_test.shape

((272, 13, 1), (31, 13, 1))

三、模型构建，编译，训练

1. 模型构建

import tensorflow
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, LSTM, SimpleRNN

model = Sequential()
model.add(SimpleRNN(200,input_shape = (13,1), activation = 'relu'))
model.add(Dense(100, activation = 'relu'))
model.add(Dense(1, activation = 'sigmoid'))
model.summary()

Model: "sequential_3"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 simple_rnn_3 (SimpleRNN)    (None, 200)               40400     
                                                                 
 dense_6 (Dense)             (None, 100)               20100     
                                                                 
 dense_7 (Dense)             (None, 1)                 101       
                                                                 
=================================================================
Total params: 60601 (236.72 KB)
Trainable params: 60601 (236.72 KB)
Non-trainable params: 0 (0.00 Byte)
_________________________________________________________________

2. 模型编译

opt = tf.keras.optimizers.Adam(learning_rate = 1e-4)

model.compile(loss = 'binary_crossentropy',
              optimizer = opt,
              metrics = 'accuracy')

3. 模型训练

epochs = 60

history = model.fit(X_train,y_train,
                    epochs = epochs,
                    batch_size = 128,
                    validation_data = (X_test,y_test),
                    verbose = 1)

Epoch 1/60
3/3 [==============================] - 1s 106ms/step - loss: 0.7025 - accuracy: 0.3382 - val_loss: 0.6939 - val_accuracy: 0.3548
Epoch 2/60
3/3 [==============================] - 0s 23ms/step - loss: 0.6929 - accuracy: 0.5809 - val_loss: 0.6805 - val_accuracy: 0.7097
Epoch 3/60
3/3 [==============================] - 0s 22ms/step - loss: 0.6839 - accuracy: 0.6875 - val_loss: 0.6674 - val_accuracy: 0.8710
Epoch 4/60
3/3 [==============================] - 0s 24ms/step - loss: 0.6752 - accuracy: 0.7279 - val_loss: 0.6546 - val_accuracy: 0.8710
Epoch 5/60
3/3 [==============================] - 0s 24ms/step - loss: 0.6669 - accuracy: 0.7390 - val_loss: 0.6420 - val_accuracy: 0.8710
Epoch 6/60
3/3 [==============================] - 0s 23ms/step - loss: 0.6588 - accuracy: 0.7500 - val_loss: 0.6301 - val_accuracy: 0.8710
Epoch 7/60
3/3 [==============================] - 0s 21ms/step - loss: 0.6508 - accuracy: 0.7610 - val_loss: 0.6184 - val_accuracy: 0.8710
Epoch 8/60
3/3 [==============================] - 0s 24ms/step - loss: 0.6426 - accuracy: 0.7610 - val_loss: 0.6065 - val_accuracy: 0.8710
Epoch 9/60
3/3 [==============================] - 0s 23ms/step - loss: 0.6340 - accuracy: 0.7610 - val_loss: 0.5944 - val_accuracy: 0.8710
Epoch 10/60
3/3 [==============================] - 0s 23ms/step - loss: 0.6250 - accuracy: 0.7647 - val_loss: 0.5814 - val_accuracy: 0.8710
Epoch 11/60
3/3 [==============================] - 0s 23ms/step - loss: 0.6160 - accuracy: 0.7721 - val_loss: 0.5676 - val_accuracy: 0.8710
Epoch 12/60
3/3 [==============================] - 0s 25ms/step - loss: 0.6057 - accuracy: 0.7794 - val_loss: 0.5527 - val_accuracy: 0.8710
Epoch 13/60
3/3 [==============================] - 0s 24ms/step - loss: 0.5952 - accuracy: 0.7978 - val_loss: 0.5368 - val_accuracy: 0.8710
Epoch 14/60
3/3 [==============================] - 0s 25ms/step - loss: 0.5836 - accuracy: 0.8051 - val_loss: 0.5193 - val_accuracy: 0.9032
Epoch 15/60
3/3 [==============================] - 0s 27ms/step - loss: 0.5706 - accuracy: 0.7978 - val_loss: 0.4999 - val_accuracy: 0.9032
Epoch 16/60
3/3 [==============================] - 0s 24ms/step - loss: 0.5576 - accuracy: 0.8015 - val_loss: 0.4784 - val_accuracy: 0.9032
Epoch 17/60
3/3 [==============================] - 0s 24ms/step - loss: 0.5424 - accuracy: 0.8015 - val_loss: 0.4550 - val_accuracy: 0.9032
Epoch 18/60
3/3 [==============================] - 0s 25ms/step - loss: 0.5264 - accuracy: 0.8088 - val_loss: 0.4306 - val_accuracy: 0.8710
Epoch 19/60
3/3 [==============================] - 0s 25ms/step - loss: 0.5103 - accuracy: 0.8199 - val_loss: 0.4056 - val_accuracy: 0.8710
Epoch 20/60
3/3 [==============================] - 0s 25ms/step - loss: 0.4935 - accuracy: 0.8088 - val_loss: 0.3815 - val_accuracy: 0.8710
Epoch 21/60
3/3 [==============================] - 0s 25ms/step - loss: 0.4771 - accuracy: 0.8088 - val_loss: 0.3569 - val_accuracy: 0.8710
Epoch 22/60
3/3 [==============================] - 0s 26ms/step - loss: 0.4627 - accuracy: 0.8125 - val_loss: 0.3348 - val_accuracy: 0.8710
Epoch 23/60
3/3 [==============================] - 0s 28ms/step - loss: 0.4498 - accuracy: 0.8199 - val_loss: 0.3159 - val_accuracy: 0.8710
Epoch 24/60
3/3 [==============================] - 0s 28ms/step - loss: 0.4388 - accuracy: 0.7941 - val_loss: 0.3001 - val_accuracy: 0.8710
Epoch 25/60
3/3 [==============================] - 0s 26ms/step - loss: 0.4325 - accuracy: 0.7904 - val_loss: 0.2914 - val_accuracy: 0.8065
 ...
Epoch 59/60
3/3 [==============================] - 0s 26ms/step - loss: 0.3221 - accuracy: 0.8640 - val_loss: 0.2794 - val_accuracy: 0.9032
Epoch 60/60
3/3 [==============================] - 0s 27ms/step - loss: 0.3244 - accuracy: 0.8640 - val_loss: 0.2833 - val_accuracy: 0.9032

四、模型评估

import matplotlib.pyplot as plt
acc = history.history['accuracy']
val_acc = history.history['val_accuracy']
loss = history.history['loss']
val_loss = history.history['val_loss']

epochs_range = range(epochs)

plt.figure(figsize=(14, 4))
plt.subplot(1, 2, 1)

plt.plot(epochs_range, acc, label='Training Accuracy')
plt.plot(epochs_range, val_acc, label='Test Accuracy')
plt.legend(loc='lower right')
plt.title('Training and Validation Accuracy')

plt.subplot(1, 2, 2)
plt.plot(epochs_range, loss, label='Training Loss')
plt.plot(epochs_range, val_loss, label='Test Loss')
plt.legend(loc='upper right')
plt.title('Training and Validation Loss')
plt.show()

scores = model.evaluate(X_test,y_test,verbose=0)
print("%s: %.2f%%" % (model.metrics_names[1],scores[1]*100))

accuracy: 90.32%

当训练轮次过高后，测试集准确率会下降，可能达到了过拟合，所以修改epoch在60时，达到了比较好的性能。