FICO Awarded 13 New Patents for Responsible AI, Fraud and Decision Management

Silicon Valley analytic software firm FICO has been awarded 13 new patents related to fraud, artificial intelligence (AI) and decision management. In total, FICO currently holds 198 US and foreign patents and 103 pending patent applications.

Read More: Simple Announces Launch of Tax Refund Feature to Automate Savings for Customers

Five of the patents are for ground-breaking analytic technology used by FICO’s industry-leading FICO® Falcon® Platform for fraud management:

• Cloud Process for Rapid Data Investigation and Data Integrity Analysis, a patent now granted in France, Germany and the UK, relates to the FICO® Falcon® Intelligence Network and FICO® Falcon® Platform CloudEdition. Superior fraud detection is achieved by analyzing an abundance of transactional data in order to effectively understand behavior and assess risk, at an individual level. This new patent helps improve the speed and quality of the data being analyzed.
• Real Time Autonomous Archetype Outlier Analytics covers a method and system for identifying subsets of transactions. This new patent identifies transactions that generate anomalies in behavioral archetypes and alerting in a score. This innovation is utilized in FICO financial crime models to detect fraud and money laundering with pinpoint accuracy. FICO® Siron® AML and FICO® Falcon® Platform use these models.
• False Positive Reduction in Abnormality Detection System Models. This invention achieves false positive reduction in transactions with recurrences in a consumer’s previous transaction history. This innovation further identifies non-fraud transactions that occur during a fraud episode, thereby improving the customer experience. This is an important part of our approach to improve in-stream analytics with the FICO® Falcon® Platform.
• Auto Adaptive Anomaly Detection System for Streams identifies click fraud in connection with Internet search engine advertising.
• Adaptive Fraud Detection relates to a real-time adaptive system for fraud detection implementing automatic binning, feature selection, adaptive statistical models, and score blending. The adaptive technology overlays core machine learning technology to adjust to near-term fraud and non-fraud activity occurring between model retrains. This technology is a critical component of machine learning models within the FICO® Falcon® Platform.

Read More: Kensho Collaborates with NVIDIA to Advance Automatic Speech Recognition