The project: Prediction of a heart disease

Description of the project

The goal of the project is to predict heart diseases based on the given data. The target variable is 'cardio'. The evalution metric for the project is ROC-AUC.

The given dataset gives the following parameters of the objects:

• id;
• age;
• gender;
• height;
• weight;
• ap_hi;
• ap_lo;
• cholesterol;
• gluc;
• smoke;
• alco;
• active.

Plan of the project:

1. Import of Modules & Files Opening
2. Data Preprocessing and Exploratory Data Analysis (EDA)
3. Development of ML-models
4. Conclusion

Fulfilled tasks:

1. Created a ML-model (RandomForestClassifier) for the regression task, which determines the presence of a heart diseases. ROC-AUC of the model: 0.739. The hyperparameters of the model are the following: criterion = 'entropy', n_estimators = 350, max_depth = 10.
2. Created a web-application using which a person can find the risk of a heart disease (that's a model. In case of anything, refer to a specialist).

Used libraries:

matplotlib, numpy. pandas, phik, pickle, seaborn, sklearn