This web application determines if a webpage is a phishing page or not using the Random Forest Classification Algorithm.
The web application is hosted at https://phishing-domain-detection.herokuapp.com/
- Four different machine learning algorithms were used to train our model (Random Forest, Perceptron, Linear Regression, and CART).
- The table below shows the accuracy of the trained models for five different testing-training data splits.
| Algorithm | 50-50 Split | 60-40 Split | 70-30 Split | 80-20 Split | 90-10 Split |
|---|---|---|---|---|---|
| Logistic Regression | 78.09% | 78.02% | 78.35% | 78.43% | 80.83% |
| Random Forest | 85.11% | 84.58% | 85.71% | 85.48% | 88.52% |
| Perceptron | 68.90% | 74.36% | 72.41% | 68.11% | 73.33% |
| C.A.R.T. | 84.53% | 84.35% | 85.26% | 84.71% | 87.70% |
- The model trained using the Random Forest Classification Algorithm was selected as it consistently provided the highest accuracy (between 85% and 88%)