Heart Failure Clinical Data Analysis using Python (Machine Learning)

By Debolina Poddar

This project is based on machine learning using Python language.We apply several machine learning classifiers to predict the patients survival.

This project is based on machine learning using Python language. We apply several machine learning classifiers to predict the patient's survival.

The heart Failure Clinical dataset contains the medical records of 299 patients who had heart failure. The dataset contains 11 clinical features, the follow-up period, and the label DEATH_EVENT that indicates whether or not the patient has died.

In this project we take a dataset then we do EDA and understand the data set then applying some preprocessing and apply a classification algorithm.
We can find some features strictly related to medical aspects like levels of enzymes, sodium, creatinine, and platelets in the blood and others that are more common like age, sex, or smoking. In this project, we take the first dataset and then we do EDA to understand the dataset and applying to preprocess and apply some classification algorithms like Decision Tree Classifier, KNeighborsClassifier, Random Forest classifier, SVM, Logistic Regression.

Main libraries are

Numpy: standard library for math operations.

Pandas: Used to manipulate data inside data frames and for basic computations.

Sklearn: used to apply different ML models.

Pyplot: To plot Visualizations

Seaborn: built on top of Pyplot

STEP 1:

import all libraries

STEP 2

Import CSV file

STEP 3

EDA

STEP 4

Preprocessing

 

building model

i)SVM

Accuracy: 82.66666666666667

ii)Decision Tree Classifier

Accuracy: 78.66666666666666

iii)KNeighborsClassifier

Accuracy: 77.33333333333333

iv)Random Forest classifier

Accuracy: 86.66666666666667

v)Logistic Regression.

Accuracy: 86.66666666666667