Phishing Websites
This experiments consist of a ML model that predicts whether a website is a phishing one or not.
This experiments consist of a ML model that predicts whether a website is a phishing one or not.
**Approach:**
1. Obtained Data set from https://archive.ics.uci.edu/ml/datasets/Phishing+Websites
2. Convert the data set into CSV format.
3. Select the required/important features
![Features importance][1]
[1]: https://drive.google.com/open?id=14smjsiRFbJx2d3SgXfOagxUEzsukrZux
4. Split the data set into training and testing data.(Training data-80% ,Testing Data-20%)
5. Train the model with training data using Two-class Decision tree Algorithm.(contains 30 trees).
6. use test data for prediction .
7. Find accuracy ,confusion matrix and other metrics and analyze the model.
Current model gives ~96% accuracy using split-sample technique
