Silicon Valley analytic software firm FICO has been awarded eight new patents related to cybersecurity and advanced analytics. FICO now holds 173 US and foreign patents, and has 91 pending patent applications.
In the area of cybersecurity, FICO’s new patent for “Cyber Security Adaptive Analytics Threat Monitoring System and Method" uses self-learning artificial intelligence models and novel machine learning adaptive behavioural analytics to detect malware. Network flow protocol messages are monitored from client computers to servers, in order to determine a risk that those client computers are behaving abnormally, which may be a result of malware, cybercriminal take-over, or botnet. Real-time entity profiles are generated for multiple entities, including client computers.
“Using the real-time entity profiles and advanced adaptive behavioural analytics, we can determine the risk that a computer is infected by malware or taken over by a cybercriminal,” said Dr. Scott Zoldi, FICO chief analytic officer and one of three inventors of the patent. “Scores are then generated in real time, representing probabilities that client computers are at risk of active attack.”
This invention is used in FICO® Cyber Security Analytics.
The additional patents cover inventions that:
- Reduce the complexity of neural network sensitivity analysis. This invention greatly reduces complexity of the calculations required to generate outputs of a neural network model, a form of artificial intelligence. This technology is implemented in FICO® Model Builder.
- Enable scoring models to select the variables with the highest predictive power (divergence) and the lowest overlap, when scoring a transaction. This invention is also used in FICO® Model Builder.
- Turn complex score models into programming code. This invention can be used to accelerate the transformation of a scoring model into a program using a concurrent, class-based, object-oriented computer programming language, such as Java, C and COBOL. This technology is integrated in FICO® Blaze Advisor® decision rules management system.
- Reconcile records from multiple data sources, eliminating discrepancies that would lead to revenue leakage or over-billing of a service.
- Use a relational logic management system to reconcile multiple interrelated rule sets in a workflow model of a business process. This allows business analysts and professionals to express the policies, tasks, rules and objectives in a format that allows interoperability with existing business procedures and automated systems without cumbersome coding requirements. This invention, granted a patent in China, is used in FICO® Blaze Advisor® decision rules management system.
- Improve selective advertising using credential information submitted by a user.
- Execute secure commercial transactions online through a system that intercepts and verifies, against consumer provided inputs, Secure Socket Layer (SSL) communications from normal web browser usage.
“Our investments in artificial intelligence, advanced analytics and decision science are backed by the strongest research team in our field,” said Dr. Stuart Wells, FICO’s chief product and technology officer. “We are bringing analytics and software breakthroughs to the commercial market at the fastest rate in our 61-year history.”