Authentication Risk Score Distribution Deviation Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Adaptive authentication systems face challenges in managing deviations between expected and actual risk scores, leading to insufficient manpower in call centers due to discrepancies between predicted and actual high-risk transactions, which can result in inadequate handling of fraudulent activities.
Innovation Solution
A computer-implemented method and apparatus that compare expected and actual risk score distributions using a Kolmogorov-Smirnov statistic to identify deviations, outputting alerts based on predefined policies, allowing for dynamic normalization and resource allocation to match the actual risk level.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the risk engine deploys a new model that increases the amount of transactions receiving a high risk score, then the detection of fraudulent transactions is improved, but the service provider experiences insufficient manpower in the call center to handle the increased volume of high-risk transactions
Solution Approach 1:
The system performs preliminary normalization of risk scores to predict the expected number of high-risk transactions before actually processing them. This allows the service provider to proactively allocate call center manpower resources to match the anticipated volume of fraudulent transactions, preventing resource shortages when fraud detection is intensified
Solution Approach 2:
The system continuously compares actual high-risk transaction volumes against expected volumes and generates alerts when deviations occur. This feedback mechanism enables dynamic adjustment of call center resource allocation, ensuring that manpower availability matches the actual fraud detection workload in real-time
2Ease of operation
If the risk engine generates normalized risk scores to maintain a constant percentage of transactions in a high risk band, then resource planning becomes easier, but significant deviations between expected and actual high-risk transactions can occur, leading to inadequate resource allocation
Solution Approach 1:
The system implements a feedback mechanism that continuously monitors the deviation between expected and actual high-risk transaction volumes. When the actual volume significantly deviates from the expected volume, the system generates alerts that trigger dynamic resource reallocation, ensuring that resource planning remains accurate even when fraud patterns change
Solution Approach 2:
The system transitions from static resource planning based on normalized risk scores to dynamic resource allocation that adapts to actual fraud conditions. By continuously comparing expected versus actual high-risk transaction volumes and generating real-time alerts, the system enables flexible resource adjustment to match actual operational needs
Data Source
AI summary
There are disclosed techniques for use in authentication. In one embodiment, the techniques comprise generating first and second distributions. The first distribution relating to risk scores expected to be produced by an authentication system in connection with requests to access a computerized resource. The expected risk scores are based on a normalization process configured to produce risk scores by normalizing raw risk scores in connection with requests. The second distribution relates to risk scores actually produced by the authentication system in connection with requests. The actual risk scores include risk scores normalized by the normalization process. The techniques also comprise comparing the first and second distributions by determining a Kolmogorov-Smirnov distance between the respective distributions. The techniques also comprise initiating, based on the comparison, a failover of the normalization process to a new normalization process for use by the authentication system.