Fair Isaac. has filed a patent for a system that enhances the security of computer-based artificial intelligence by monitoring transactions, detecting adversarial latent features, and blocking potentially harmful transactions. The system aims to improve the accuracy and reliability of machine learning decision models. GlobalData’s report on Fair Isaac gives a 360-degree view of the company including its patenting strategy. Buy the report here.
Smarter leaders trust GlobalData
According to GlobalData’s company profile on Fair Isaac, AI for workflow management was a key innovation area identified from patents. Fair Isaac's grant share as of January 2024 was 52%. Grant share is based on the ratio of number of grants to total number of patents.
Improving security of computer-implemented artificial intelligence by monitoring transactions
A computer-implemented artificial intelligence system described in a filed patent (Publication Number: US20240039934A1) aims to enhance security by monitoring transactions received by a machine learning decision model. The system involves receiving scores for transactions, identifying transactions belonging to specific classes based on score thresholds and occurrence likelihood, and detecting adversarial latent features. These features are then used to block potentially malicious transactions and generate top reasons for system-level scores above a threshold. The system utilizes moving average features, quantile estimation processes, and neural network models to improve security measures and prevent adversarial attacks.
Furthermore, the patent outlines a method for improving the security of the artificial intelligence system by monitoring transactions, identifying likely adversarial transactions based on low scores and occurrence likelihood, and training an adversary detection model. The system generates a corpus of transactions, calculates scores, and distinguishes likely adversarial transactions from others. The adversary detection model, which can be a neural network or a multi-layered self-calibrating model, is trained using attributes of the transactions in the corpus. The training process involves minimizing a cost function to separate actual and predicted tag values, ensuring accurate identification of likely adversarial transactions. By utilizing advanced machine learning techniques and models, the system aims to enhance security measures and protect against potential adversarial attacks.
To know more about GlobalData’s detailed insights on Fair Isaac, buy the report here.
Premium Insights
From
The gold standard of business intelligence.
Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.
GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.
GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.
