Identifying patients with heart disease
Source: https://www.kaggle.com/datasets/bharath011/heart-disease-classification-dataset
Used in accordance with the Attribution 4.0 International license.
Import libraries¶
import pandas as pd
import seaborn as sns
from sklearn import svm
from sklearn.preprocessing import LabelEncoder
from sklearn.model_selection import train_test_split
data = pd.read_csv('Heart Attack.csv')
data
| age | gender | impluse | pressurehight | pressurelow | glucose | kcm | troponin | class | |
|---|---|---|---|---|---|---|---|---|---|
| 0 | 64 | 1 | 66 | 160 | 83 | 160.0 | 1.80 | 0.012 | negative |
| 1 | 21 | 1 | 94 | 98 | 46 | 296.0 | 6.75 | 1.060 | positive |
| 2 | 55 | 1 | 64 | 160 | 77 | 270.0 | 1.99 | 0.003 | negative |
| 3 | 64 | 1 | 70 | 120 | 55 | 270.0 | 13.87 | 0.122 | positive |
| 4 | 55 | 1 | 64 | 112 | 65 | 300.0 | 1.08 | 0.003 | negative |
| ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
| 1314 | 44 | 1 | 94 | 122 | 67 | 204.0 | 1.63 | 0.006 | negative |
| 1315 | 66 | 1 | 84 | 125 | 55 | 149.0 | 1.33 | 0.172 | positive |
| 1316 | 45 | 1 | 85 | 168 | 104 | 96.0 | 1.24 | 4.250 | positive |
| 1317 | 54 | 1 | 58 | 117 | 68 | 443.0 | 5.80 | 0.359 | positive |
| 1318 | 51 | 1 | 94 | 157 | 79 | 134.0 | 50.89 | 1.770 | positive |
1319 rows × 9 columns
Data preparation¶
le = LabelEncoder()
le.fit(data['class'])
data['class'] = le.transform(data['class'])
data.corr()['class']
age 0.238097 gender 0.094432 impluse 0.006920 pressurehight -0.020825 pressurelow -0.009659 glucose -0.033059 kcm 0.217720 troponin 0.229376 class 1.000000 Name: class, dtype: float64
Data analysis¶
data.plot.hist(y="kcm")
data.plot.hist(y="troponin")
<Axes: ylabel='Frequency'>
Training the model¶
X = data.drop('class', axis=1)
y = data['class']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=1)
svm_model = svm.SVC()
svm_model.fit(X_train, y_train)
# Make predictions
y_pred = svm_model.predict(X_test)
y_pred
array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0,
1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
0, 0, 1, 1, 1, 1, 1, 0, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1,
1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0,
1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1])
from sklearn.metrics import mean_absolute_error
error = mean_absolute_error(y_test, y_pred)
error
0.3446969696969697
data.describe()['class']
count 1319.000000 mean 0.614102 std 0.486991 min 0.000000 25% 0.000000 50% 1.000000 75% 1.000000 max 1.000000 Name: class, dtype: float64
Writer's note: the error is below the STD indicating relative good performance.
from sklearn.metrics import accuracy_score, f1_score, recall_score, precision_score, roc_curve, roc_auc_score
import matplotlib.pyplot as plt
accuracy = accuracy_score(y_test, y_pred)
f1 = f1_score(y_test, y_pred)
recall = recall_score(y_test, y_pred)
precision = precision_score(y_test, y_pred)
print("Accuracy is ", accuracy)
print("F1-score is ", f1)
print("Recall is ", recall)
print("Precision is ", precision)
Accuracy is 0.6553030303030303 F1-score is 0.7707808564231738 Recall is 0.9683544303797469 Precision is 0.6401673640167364
from sklearn.metrics import confusion_matrix
cm = confusion_matrix(y_test, y_pred)
import matplotlib.pyplot as plt
import seaborn as sns
plt.figure(figsize=(8, 6))
sns.heatmap(cm, annot=True, fmt='d', cmap='Blues', cbar=False)
plt.xlabel('Predicted')
plt.ylabel('True')
plt.title('Confusion Matrix')
plt.show()
Conclusion¶
From the common performance metrics as well as the confusion matrix we can state that the overall performance of the support vector machine model on this heart disease dataset is good. Considering the context, however, which is to classify patients with heart disease correctly, from the results of the confusion matrix it becomes clear that the number of false positives is relatively low. This further supports the claim that the model its performance is good. Through the results it also becomes clear that the model reaches a lot of false positive cases. This will lead to more work for hospital staff, nurses and doctors. The model appears to have trouble with correctly identifying patients that do not have heart disease. Consequently, further research is warranted to explore this fault and to enhance the model its reliability and robustness.