Heart disease prediction using Keras

By ADIL SHAIKH

This project is meant for detecting the is the person has a heart disease given certain parameters as input.

This project provides an insight into the Keras framework by using a shallow neural network for predicting if the person has heart disease. The dataset has been taken from Kaggle and the accuracy of the model will increase by increasing the amount of data used in training. This project can also be further developed to predict a heart attack even before it happens and can be employed in wearable technology. This project uses a shallow neural network but the number of layers and neurons can be increased or decreased depending on the size of the dataset.

This project uses the following libraries:

DATASET: Heart disease dataset

REQUIREMENTS:

1. Tensorflow 

2. Python 

3. Keras 

4. pandas

5. Seaborn

6. Matplotlib

 

Descriptors in the dataset used to predict heart disease:

1. age - age in years

2. sex - (1 = male; 0 = female)

3. cp - chest pain 

4. trestbps - resting blood pressure (in mm Hg on admission to the hospital)

5. chol - serum cholesterol in mg/dl

6. fbs - (fasting blood sugar > 120 mg/dl) (1 = true; 0 = false)

7. restecg - resting electrocardiographic results

8. thalach - maximum heart rate achieved

9. exang - exercise-induced angina (1 = yes; 0 = no)

10. oldpeak - ST depression induced by exercise relative to rest

11. slope - the slope of the peak exercise ST segment

12. ca - number of major vessels (0-3) colored by flourosopy

13. thal - thallium stress result

14. target - have a disease or not (1=yes, 0=no) (= is the predicted feature)

 

USAGE :

You must have Jupiter notebook installed in your pc.

There are 2 ways to use the given zip file:

1) Directly use the Keras model saved 'heart_disease_predict.h5' in your own personal prediction model. This is present in the zip folder

2) Train using your own dataset by changing the dataset file location at "df = pd.read_csv("/home/adil/python3_dir/heart_disease_prediction/datasets_heart.csv") " to suit your purpose. This will also help you visualize the data present in your own dataset. Run the file "heart_disease_prediction_using_keras.ipynb. The code has been explained inside the file.