Analytics software firm FICO receives 12 patents

‘Efficient Association of Related Entities’ – Relating to identification and tracking of associations between related entities in a distributed computing environment to readily determine associations between entities known to be involved in fraud.

‘Blockchain for Data and Model Governance’ – Covering the concept of a shared ledger that is used to track end-to-end provenance of the development, operationalization, and monitoring of a machine learning models in an immutable manner. The solution and underlying methodology enforce the use of a corporate model development standard to ensure Responsible AI practices are abided by organizations.

‘Providing Insights About Dynamic Machine Learning Models’ – Encompassing the use of visual indicators of features and training data, along with global or local explanations, to provide a broader summarization of a machine learning models behaviour.

‘Managing Missing Values in Datasets for Machine Learning Models’ — Addressing missing value data, imputing values, and encoding reasons for why data may be missing. This invention is for use in the development of analytic models in environments where data quality is suspect.

‘Facial Recognition for User Authentication’ – Covering the use of facial liveness recognition for the purpose of providing secure access to electronic content or systems. The facial features of the user are continually scanned to detect facial movements indicative of the user’s liveness, as opposed to static representations.

‘Deployment of Self-Contained Decision Logic’ – Encompassing the generation of executable code associated with the decision logic and makes it amenable to storage in a repository.

‘Devices And Methods for Efficient Execution of Rules Using Pre-Compiled Directed Acyclic Graphs’ — Pertaining to making optimal transformations of business rules into executable code for rapid execution.

‘Systems and Methods for Scalable, Adaptive, Real Time Personalized Offers Generation’ – Covering improvements to hyper-personalization offers by updating user profiles in real-time with each new transaction or event, generating an affinity score for each user based on the updated profile.

‘Density Based Confidence Measures of Neural Networks for Reliable Predictions’ – Claiming innovations in assigning confidence estimates to predictions, thus allowing practitioners to better trust the model. This invention opens new applications in which knowing the certainty of the predictions is critically important to the implementation of responsible AI.

‘Efficient Value Lookup In A Set Of Scalar Intervals’ – Encompassing improvements to sorting and comparing scalar intervals to determine overlaps and make decisions faster.

‘Entity Resolution Based on Character String Frequency Analysis’ – Addressing the generation of similarity scores between one or more pairs of character strings based on degrees of commonality or rarity, generating a similarity score.

‘Advanced Learning System for Detection and Prevention of Money Laundering’ — Covering automated detection of money laundering using hyper personalization of linked list behaviour to known risky entities. This innovation is used to create money laundering threat scores to help detect and stop money laundering.