An application that allows users to predict the risk of a patient having parkisnsons based on data from microphone recordings
-
Performed various Data Preprocessing techniques such as feature scaling and class balancing to clean and make the data ready for model building
-
Random Forest, XGBoost and Support Vector Machines (SVM) were used to model the data
-
Random Forest was the best with an accuracy of 97.44%, f1-score of 98.41% and recall of 100%
-
Model was deployed on a web application built using Django available in the
web-app/folder
Accuracy, F1-Score and Recall were the metrics used to evaluate the performance of the model
| Method | Accuracy (%) | F1-Score (%) | Recall (%) |
|---|---|---|---|
| Random Forest | 97.44 | 98.41 | 100.00 |
| XGBoost | 89.74 | 93.94 | 100.00 |
| SVM | 89.74 | 93.33 | 90.32 |
Confusion Matrix
| 0 | 1 | |
|---|---|---|
| 0 | TN | FP |
| 1 | FN | TP |
-
Data was shuffled and split with a 80/20 ratio before preprocessing to avoid data leakage
-
There was no missing values in the data, so no need for imputation
-
Feature Scaling was applied using Standardization method
-
Then the Class Imbalance was fixed from a 74:26 ratio to a 50:50 ratio
The final model with the best score was deployed on a web application built with Django with the frontend built with HTML & CSS with Boostrap 4 as the CSS Framework.
