Browse Prior Art Database

Cyber Beach Prediction Algorithm Based on Machine Learning Method (Poisson's cumulative distribution function) and Process

IP.com Disclosure Number: IPCOM000250418D
Publication Date: 2017-Jul-12
Document File: 3 page(s) / 446K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed are a method and tangible tool that help determine the likelihood of a cyber breach for a company located in a certain region or industry based on analyzing and correlating past historical breach data and predicting breach probability using Poisson's cumulative distribution function and provable statistics methods.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

1

Cyber Beach Prediction Algorithm Based on Machine Learning Method (Poissons Cumulative Distribution Function) and Process

The novel contribution is a method and tangible tool that helps determine the likelihood of a cyber breach for a company located in a certain region or industry based on analyzing and correlating past historical breach data and predicting breach probability using Poisson's cumulative distribution function and provable statistics methods. The machine learning method is used in conjunction with the analysis of security controls effectiveness as determined using International Organization for Standardization (ISO), Information Security Forum (ISF), SANS, or other security frameworks. This method and the accompanying tool bring together an algorithm for determining the breach likelihood. Currently, the breach likelihood is evaluated using a consultative approach and qualitative information only.

The solution, which is comprised of the method and the tangible tool, is based on a machine learning algorithm that can be implemented using a cognitive analytics system, a statistical inference tool, or visualization software. The idea comes from using Poisson's regression analysis and cumulative distribution function in predicting the probability of cyber breach for a company based on past historical data of breach incidents. The method analyzes the past breach patterns for a sector or country (region), and then determines the probability of the likelihood of a breach for the target client based on frequency and variety of events....