Leading digital decision platform company, FICO, today announced that it has been awarded 13 new patents related to artificial intelligence (AI), machine learning (ML), fraud and decision management platform. With the latest patents, FICO continues to help its customers digitally transform their businesses by automating key business processes and decisioning with industry-leading innovations.
With the 13 new innovations, FICO now holds 204 US and foreign patents and 85 pending patent applications. FICO has been responsible for multiple industry-changing innovations in artificial intelligence, machine learning, and other analytics methods. FICO’s rich portfolio of analytics and fraud solutions helps clients grapple with ever larger volumes and variety of data across the enterprise as well as protect businesses against the latest fraud in real time.
“FICO provides our customers the solutions they need at the moments they need them, and driving forward their success is what allows us to thrive as an organization. We have created and nourished an environment that empowers my colleagues and myself to push the boundaries and continuously drive innovations that helps customers succeed. It’s exciting to see our work receive that recognition,” said Scott Zoldi, chief analytics officer, FICO.
The patents awarded to FICO and their innovative executives include:
- “Explaining Machine Learning Models By Tracked Behavioral Latent Features” by Scott Zoldi. This invention is a system and method to explain machine learning model behavior, which can benefit not only those seeking to meet regulatory requirements when using models but also help guide users of models to assess and increase robustness associated with model governance processes. This innovation is utilized in the FICO® Falcon® Fraud Manager and FICO® Falcon® X models.
- “Fast Automatic Explanation of Scored Observations.” This patent by Gerald Fahner and Scott Zoldi relates to systems and methods for generating concise explanations of scored observations that strike good, and computationally efficient, trade-offs between rank-ordering performance and explainability of scored observations, based on a framework of partial dependence functions (PDFs), multi-layered neural networks (MNNs), and Latent Explanations Neural Network Scoring (LENNS).
- “Detection Of Compromise Of Merchants, ATMS, And Networks.” This patent by Scott Zoldi relates to the generation of compromise profiles for financial merchants and accounts based on a comparison of reported fraud data with an account profile, account transaction profile, merchant device profile, and merchant device account history profile–to quickly identify when account information has been obtained by an unauthorized third party and when. The systems and methods claimed by the patent relate to FICO offerings for point of compromise and mass compromise detection.
Also Read: Reshaping the Post-pandemic enterprise landscape with Data Science
- “System and Method for Linearizing Messages from Data Sources for Optimized High-Performance Processing in a Stream Processing System.” This innovation by Shalini Raghavan and Tom Traughber relates to the processing of data objects by a distributed stream computing system, and more specifically, the linearized processing of data objects. This technology is integrated with FICO® Decision Management Platform Streaming.
- Multi-Layered Self-Calibrating Analytics,” an invention by Scott Zoldi presents multi-layered, self-calibrating analytics for detecting fraud in transaction data without substantial historical data including limited or no outcome data. In markets where transaction history data is not widely available, this invention enables an adaptive selection and grouping of variables relating to real-time transaction data, for processing by a number of independent self-calibrating models. The outputs of these models are combined for an accurate fraud score based on anomaly detection of discovered hidden latent features.
- “Behavioral Misalignment Detection within Entity Hard Segmentation Utilizing Archetype-Clustering” by Scott Zoldi and Joe Murray. This invention is an automated way of learning archetypes which capture many aspects of entity behavior, and assigning entities to a mixture of archetypes, such that each entity is represented as a distribution across multiple archetypes. Given those representations in archetypes, anomalous behavior can be detected by finding misalignment with a plurality of entities having archetype clustering within a hard segmentation. FICO® Anti-Financial Crime Solutions uses this technology.