-
Notifications
You must be signed in to change notification settings - Fork 0
/
14_models_for_cardiovascular_disease_prediction.py
311 lines (225 loc) · 9.3 KB
/
14_models_for_cardiovascular_disease_prediction.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
# -*- coding: utf-8 -*-
"""20-models-for-cardiovascular-disease-prediction.ipynb
Automatically generated by Colab.
Original file is located at
https://colab.research.google.com/drive/1KM0bgi6fZebH1DqLsxKn2vKnRS3BDDza
"""
# !pip install --upgrade pandas-profiling
# !pip install pydantic-settings
# Commented out IPython magic to ensure Python compatibility.
!pip install --upgrade pandas-profiling
!pip install pydantic-settings
!pip install --upgrade tensorflow
!pip install scikeras
!pip install --upgrade scikit-learn
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
# %matplotlib inline
# preprocessing
from sklearn.preprocessing import LabelEncoder
from sklearn.model_selection import train_test_split, cross_val_score, GridSearchCV
# models
from sklearn.linear_model import LogisticRegression, Perceptron, RidgeClassifier, SGDClassifier
from sklearn.svm import SVC, LinearSVC
from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier, ExtraTreesClassifier
from sklearn.ensemble import BaggingClassifier, AdaBoostClassifier, VotingClassifier
from sklearn.neighbors import KNeighborsClassifier
from sklearn.naive_bayes import GaussianNB
from sklearn.tree import DecisionTreeClassifier
from sklearn import metrics
import xgboost as xgb
from xgboost import XGBClassifier
import lightgbm as lgb
from lightgbm import LGBMClassifier
# NN models
import tensorflow as tf # Import tensorflow
from tensorflow.keras.models import Sequential # Use tensorflow.keras
from tensorflow.keras.layers import Dense, Dropout
from tensorflow.keras import optimizers
from scikeras.wrappers import KerasClassifier # Import from the correct location
from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint
# model tuning
from hyperopt import STATUS_OK, Trials, fmin, hp, tpe, space_eval
# import warnings filter
from warnings import simplefilter
# ignore all future warnings
simplefilter(action='ignore', category=FutureWarning)
data = pd.read_csv("/content/cardio_train.csv", sep=";")
data.drop("id",axis=1,inplace=True)
data.drop_duplicates(inplace=True)
data["bmi"] = data["weight"] / (data["height"]/100)**2
out_filter = ((data["ap_hi"]>250) | (data["ap_lo"]>200))
data = data[~out_filter]
len(data)
out_filter2 = ((data["ap_hi"] < 0) | (data["ap_lo"] < 0))
data = data[~out_filter2]
data.head(3)
data.info()
#
!pip install --upgrade typeguard
!pip install typing_extensions
!pip install ydata-profiling
import ydata_profiling as yp
yp.ProfileReport(data)
target_name = 'cardio'
data_target = data[target_name]
data = data.drop([target_name], axis=1)
train, test, target, target_test = train_test_split(data, data_target, test_size=0.2, random_state=0)
train.head(3)
test.head(3)
train.info()
test.info()
#%% split training set to validation set
Xtrain, Xval, Ztrain, Zval = train_test_split(train, target, test_size=0.2, random_state=0)
# Logistic Regression
logreg = LogisticRegression()
logreg.fit(train, target)
acc_log = round(logreg.score(train, target) * 100, 2)
acc_log
acc_test_log = round(logreg.score(test, target_test) * 100, 2)
acc_test_log
coeff_df = pd.DataFrame(train.columns.delete(0))
coeff_df.columns = ['Feature']
coeff_df["Correlation"] = pd.Series(logreg.coef_[0])
coeff_df.sort_values(by='Correlation', ascending=False)
# Support Vector Machines
svc = SVC()
svc.fit(train, target)
acc_svc = round(svc.score(train, target) * 100, 2)
acc_svc
acc_test_svc = round(svc.score(test, target_test) * 100, 2)
acc_test_svc
# Linear SVC
linear_svc = LinearSVC(dual=False) # dual=False when n_samples > n_features.
linear_svc.fit(train, target)
acc_linear_svc = round(linear_svc.score(train, target) * 100, 2)
acc_linear_svc
acc_test_linear_svc = round(linear_svc.score(test, target_test) * 100, 2)
acc_test_linear_svc
# k-Nearest Neighbors algorithm
knn = GridSearchCV(estimator=KNeighborsClassifier(), param_grid={'n_neighbors': [2, 3]}, cv=10).fit(train, target)
acc_knn = round(knn.score(train, target) * 100, 2)
print(acc_knn, knn.best_params_)
acc_test_knn = round(knn.score(test, target_test) * 100, 2)
acc_test_knn
# Gaussian Naive Bayes
gaussian = GaussianNB()
gaussian.fit(train, target)
acc_gaussian = round(gaussian.score(train, target) * 100, 2)
acc_gaussian
acc_test_gaussian = round(gaussian.score(test, target_test) * 100, 2)
acc_test_gaussian
# Perceptron
perceptron = Perceptron()
perceptron.fit(train, target)
acc_perceptron = round(perceptron.score(train, target) * 100, 2)
acc_perceptron
acc_test_perceptron = round(perceptron.score(test, target_test) * 100, 2)
acc_test_perceptron
# Stochastic Gradient Descent
sgd = SGDClassifier()
sgd.fit(train, target)
acc_sgd = round(sgd.score(train, target) * 100, 2)
acc_sgd
acc_test_sgd = round(sgd.score(test, target_test) * 100, 2)
acc_test_sgd
# Decision Tree Classifier
decision_tree = DecisionTreeClassifier()
decision_tree.fit(train, target)
acc_decision_tree = round(decision_tree.score(train, target) * 100, 2)
acc_decision_tree
acc_test_decision_tree = round(decision_tree.score(test, target_test) * 100, 2)
acc_test_decision_tree
# Random Forest
random_forest = GridSearchCV(estimator=RandomForestClassifier(), param_grid={'n_estimators': [100, 300]}, cv=5).fit(train, target)
random_forest.fit(train, target)
acc_random_forest = round(random_forest.score(train, target) * 100, 2)
print(acc_random_forest,random_forest.best_params_)
acc_test_random_forest = round(random_forest.score(test, target_test) * 100, 2)
acc_test_random_forest
!pip install xgboost
from xgboost import XGBClassifier
# Define the parameters for the XGBoost classifier
params = {
'learning_rate': 0.1, # Example parameter, adjust as needed
'max_depth': 3, # Example parameter, adjust as needed
# Add other parameters as required
}
XGB_Classifier = XGBClassifier(**params)
XGB_Classifier.fit(train, target)
acc_XGB_Classifier = round(XGB_Classifier.score(train, target) * 100, 2)
acc_XGB_Classifier
acc_test_XGB_Classifier = round(XGB_Classifier.score(test, target_test) * 100, 2)
acc_test_XGB_Classifier
acc_test_XGB_Classifier = round(XGB_Classifier.score(test, target_test) * 100, 2)
acc_test_XGB_Classifier
fig = plt.figure(figsize = (15,15))
axes = fig.add_subplot(111)
xgb.plot_importance(XGB_Classifier,ax = axes,height =0.5)
plt.show();
plt.close()
LGB_Classifier = LGBMClassifier(**params)
LGB_Classifier.fit(train, target)
acc_LGB_Classifier = round(LGB_Classifier.score(train, target) * 100, 2)
acc_LGB_Classifier
acc_test_LGB_Classifier = round(LGB_Classifier.score(test, target_test) * 100, 2)
acc_test_LGB_Classifier
fig = plt.figure(figsize = (15,15))
axes = fig.add_subplot(111)
lgb.plot_importance(LGB_Classifier,ax = axes,height = 0.5)
plt.show();
plt.close()
# Gradient Boosting Classifier
gradient_boosting = GradientBoostingClassifier(**params)
gradient_boosting.fit(train, target)
acc_gradient_boosting = round(gradient_boosting.score(train, target) * 100, 2)
acc_gradient_boosting
acc_test_gradient_boosting = round(gradient_boosting.score(test, target_test) * 100, 2)
acc_test_gradient_boosting
# Ridge Classifier
ridge_classifier = RidgeClassifier()
ridge_classifier.fit(train, target)
acc_ridge_classifier = round(ridge_classifier.score(train, target) * 100, 2)
acc_ridge_classifier
acc_test_ridge_classifier = round(ridge_classifier.score(test, target_test) * 100, 2)
acc_test_ridge_classifier
# Bagging Classifier
bagging_classifier = BaggingClassifier()
bagging_classifier.fit(train, target)
Y_pred = bagging_classifier.predict(test).astype(int)
acc_bagging_classifier = round(bagging_classifier.score(train, target) * 100, 2)
acc_bagging_classifier
acc_test_bagging_classifier = round(bagging_classifier.score(test, target_test) * 100, 2)
acc_test_bagging_classifier
models = pd.DataFrame({
'Model': ['Logistic Regression', 'Support Vector Machines', 'Linear SVC', 'k-Nearest Neighbors', 'Naive Bayes',
'Perceptron', 'Stochastic Gradient Decent',
'Decision Tree Classifier', 'Random Forest', 'XGBClassifier', 'LGBMClassifier',
'GradientBoostingClassifier', 'RidgeClassifier', 'BaggingClassifier'],
'Score_train': [acc_log, acc_svc, acc_linear_svc, acc_knn, acc_gaussian,
acc_perceptron, acc_sgd,
acc_decision_tree, acc_random_forest, acc_XGB_Classifier, acc_LGB_Classifier,
acc_gradient_boosting, acc_ridge_classifier, acc_bagging_classifier],
'Score_test': [acc_test_log, acc_test_svc, acc_test_linear_svc, acc_test_knn, acc_test_gaussian,
acc_test_perceptron, acc_test_sgd,
acc_test_decision_tree, acc_test_random_forest, acc_test_XGB_Classifier, acc_test_LGB_Classifier,
acc_test_gradient_boosting, acc_test_ridge_classifier, acc_test_bagging_classifier]
})
models.sort_values(by=['Score_train', 'Score_test'], ascending=False)
models.sort_values(by=['Score_test', 'Score_train'], ascending=False)
models['Score_diff'] = abs(models['Score_train'] - models['Score_test'])
models.sort_values(by=['Score_diff'], ascending=True)
# Plot
plt.figure(figsize=[25,6])
xx = models['Model']
plt.tick_params(labelsize=14)
plt.plot(xx, models['Score_train'], label = 'Score_train')
plt.plot(xx, models['Score_test'], label = 'Score_test')
plt.legend()
plt.title('Score of 14 popular models for train and test datasets')
plt.xlabel('Models')
plt.ylabel('Score, %')
plt.xticks(xx, rotation='vertical')
plt.savefig('graph.png')
plt.show()