KROTON Suspect Miner
Active Learning in Suspicious Activity Detection
H3M KROTON Suspect Miner addresses the major problems related to legacy systems with its unique machine learning approach utilizing active learning.
Legacy Problem: High number of false positives resulting in excessive workload.
KROTON Solution: Eliminate 2/3 of false positives with machine learning.
Legacy Problem: Detecting less than 0.2% of suspicious activity.
KROTON Solution: Detecting 1000s of suspicious activity, which is otherwise invisible to rule-based systems. (with 98.4% true-positive)
Legacy Problem: Limited labelled data set for building a useful machine learning model.
KROTON Solution: Active Learning to incorporate analyst know-how with AI capability, and pro-active labelling to help the machines.
Legacy Problem: Rules not covering emerging laundering methods.
KROTON Solution: Active Learning routinely discovers emerging laundering methods.
Legacy Problem: The increasing backlog of alerts due to excessive alert generation.
KROTON Solution: Automatic filtering of backlog items using machine learning.
KROTON Product References & Press Releases
Industry-specific solutions
Check out our applications of the KROTON product line as a solution in different verticals:
KROTON Product Suite
H3M KROTON modules work hand-in-hand with legacy rule-based compliance systems and upgrade them with the help of Artificial Intelligence. H3M flips the paradigm. Most platforms help firms comply. Crime detection is a by-product. H3M is different. Designed from the ground up to uncover hidden links to detect and prevent crime, better compliance, enhanced regulatory relations, competitive advantage, and brand value are great by-products. The KROTON product suite includes 4 AI-based modules which powerfully combine compliance know-how with data science: