I recently worked on a project to develop machine learning models for early detection of heart diseases. The dataset I used contained 12 attributes of patients, including their age, sex, resting blood pressure, cholesterol level, and whether they have heart disease.

I first performed exploratory data analysis and visualization to gain insights into the data. I found that there was a higher proportion of males in the dataset, and that the average age of patients with heart disease was higher than those without. I also found that there was a strong correlation between cholesterol level and heart disease, and that the presence of exercise-induced angina was also a strong predictor of heart disease.

I then preprocessed the data and scaled the numerical features. I used a variety of machine learning models to train on the data, including logistic regression, decision trees, random forests, and support vector machines. I evaluated the performance of each model using the accuracy, cross-validation score, and ROC AUC score.

The best performing model was the logistic regression model, which achieved an accuracy of 87.5%, a cross-validation score of 91.12%, and a ROC AUC score of 0.87. I also evaluated the model’s performance using a confusion matrix and classification report. The confusion matrix showed that the model had a low false positive rate, which means that it was good at avoiding predicting that a patient had heart disease when they did not. The classification report showed that the model had a high precision and recall score, which means that it was good at both predicting true positives and avoiding false negatives.

Overall, I was satisfied with the performance of the logistic regression model. I believe that this model can be used to develop a system for early detection of heart diseases. In the future, I would like to explore other machine learning models and evaluate their performance on this dataset.

I hope this post has been informative and interesting. I am always happy to discuss my work with others, so please feel free to reach out if you have any questions.

Find the project here: https://github.com/Hikky99/zahamma