Heart Failure Prediction

Introduction

Cardiovascular diseases (CVDs) are the number 1 cause of death globally, taking an estimated 17.9 million lives each year, which accounts for 31% of all deaths worlwide. Heart failure is a common event caused by CVDs and this dataset contains 12 features that can be used to predict mortality by heart failure.

Most cardiovascular diseases can be prevented by addressing behavioural risk factors such as tobacco use, unhealthy diet and obesity, physical inactivity and harmful use of alcohol using population-wide strategies.

People with cardiovascular disease or who are at high cardiovascular risk (due to the presence of one or more risk factors such as hypertension, diabetes, hyperlipidaemia or already established disease) need early detection and management wherein a machine learning model can be of great help.

Ensemble Classifier Model

Stacking is an ensemble learning technique to combine multiple classification models via a meta-classifier. The individual classification models are trained based on the complete training set; then, the meta-classifier is fitted based on the outputs -- meta-features -- of the individual classification models in the ensemble. The meta-classifier can either be trained on the predicted class labels or probabilities from the ensemble. Here, in my case I have considered SVC, RF, GuassianNB combined their probabilities and used Logistic regression model as meta classifier. To measure the performance of the model I majorly considered to go on with AUC metric. To figure the how well the model is classified I have used precision and recall and F1 score.

Significant Drivers

Model Explainability

Requirement

• OS - Ubuntu 18.04
• Miniconda for linux
• Python Version 3.7.10
• PIP version 21.0.1

Instructions

Download miniconda using the following command:

wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
sh Miniconda3-latest-Linux-x86_64.sh
source .bashrc

To create conda environment to support this use-case

conda env create -f environment.yml

To activate the conda environment

conda activate Assignment

change directory to source directory

cd source/

To train an ensemble model run, open command prompt and please run the below command,

python main.py

To start the server on localhost, open commad prompt and please run the below command,

uvicorn api:app --reload

open internet browser and get into localhost server 'http://127.0.0.1:8000/docs'

Please click 'POST' menu

Please press 'Try it out' button on the right side of the bar
please paste the below sample json input in the request body

{ "age": 75, "anaemia": 0, "creatinine_phosphokinase": 582, "diabetes": 0, "ejection_fraction": 20, "high_blood_pressure": 1, "platelets": 265000.00, "serum_creatinine": 1.9, "serum_sodium": 130, "sex": 1, "smoking": 0, "time": 4 }

Finally press Execute bar.

You ca find the prediction result in Response Body as
{ "Prediction": "1" }

Curl command

To predict using curl command, open commad prompt and please run the below command,

curl -X 'POST' 'http://127.0.0.1:8000/predict' -H 'accept: application/json' -H 'Content-Type: application/json' -d '{ "age": 75, "anaemia": 0, "creatinine_phosphokinase": 582, "diabetes": 0, "ejection_fraction": 20, "high_blood_pressure": 1, "platelets": 265000.00, "serum_creatinine": 1.9, "serum_sodium": 130, "sex": 1, "smoking": 0, "time": 4 }'

Pytest

Please run 'pytest' in the command prompt for basic test like incorrect url, data, and header