Development and Deployment of a Machine Learning Model for Automatic Heart Failure Prediction

In the current era, cardiovascular diseases mainly heart failure (HF) are one of the decisive factors which lead to a high mortality rate. The main concern of this work is to deploy automation in detecting heart failure using different Machine Learning models and to create a graphical unit interface (web application). The chosen four classifier models are Logistic Regression, Decision Tree, K-nearest neighbors algorithm, and Random Forest algorithm and they were examined against the publicly available heart failure clinical record dataset from Kaggle. GridSearchCV has been utilized for fine-tuning the hyperparameters. A comparative analysis of the chosen 4 ML (Machine learning) models was also carried out and it states that the Logistic regression model has achieved the highest precision and F1 score of 89% and 92% respectively. From this work, it is apparent that heart disease prediction and diagnosis which claims to be an immense task in the medical field was facilitated by advances in Machine Learning techniques.