Cyber Insurance is a newer line of business for insurance companies. Being able to aggregate and visualize historic cyber breach data, and predict new cyber breaches can be the difference in underwriting a “good”(doesn’t blow up) risk and a “bad” risk (blows up).
Steps followed:
Data Collection: Performed web scraping and processed the data using Pandas librabry
Data Scrubbing and Transformation : Dealing with data imabalances and applying encoding techniques for training purposes
Training and Testing : Employed different classification Technqiues (SVM, Neural Net, Gradient Boosting) in order to cross validate the results