In this talk, Ashrith will be introducing you to the idea of using machine learning for detecting money laundering. The idea behind using ML for detecting money laundering is that the current rules-based engine have limited visibility into money movement. And as models learn the nuances of money movement, especially illegal, much better money laundering detection is possible.
Bio: Ashrith Barthur is a Security Scientist at H2O currently working on algorithms that detect anomalous behaviour in user activities, network traffic, attacks, financial fraud and global money movement. He has a PhD from Purdue University in the field of information security, specialized in Anomalous behaviour in DNS protocol.
https://www.linkedin.com/in/abarthur/
6. H2
O.ai
Machine Intelligence
1. Features are meta data (Extracted from the data)
2. They help algorithms capture information from the data.
3. Feature engineering is a form of language translation: Between raw data
and the algorithm.
9. H2
O.ai
Machine Intelligence
1. Designed Features Highlight Transactional Behaviour
2. Features Continuously Track Transactional Behaviour of an account
3. Rules Variables can only Identify Threshold Changes
10. H2
O.ai
Machine Intelligence
Alerts from rule-based
system
Alert Decision:
Not Suspicious
H2O Machine
Learning Algorithm
Alert Decision:
Suspicious
Analytical Inputs:
1. Transaction Data
2. Account Data
3. Card Data etc.
11. H2
O.ai
Machine Intelligence
1. Uses AI - artificial intelligence
2. AI with features uses a consistent and objective approach
3. Quick classification
4. Low false positive rate - tweaked based on risk appetite.
12. H2
O.ai
Machine Intelligence
Alerts from rule-based
system
Alert Decision:
Not Suspicious
H2O Machine Learning
Algorithm
Alert Decision:
Suspicious
Analytical Inputs:
1. Transaction Data
2. Account Data
3. Card Data etc.
AML Analyst
Alert decision sampling by the analyst
Algorithm tuning by analyst after alert
decision sampling
13. H2
O.ai
Machine Intelligence
1. AI model will learn and improve from the analyst’s feedback
2. The analyst has one single interface
3. Unified interface for dispersed datasets
15. H2
O.ai
Machine Intelligence
Not a suspicious
transaction
H2O Machine
Learning - Deep
Learning Algorithm
Suspicious
Transaction
Transaction Data
Alert Data
Card Data
Account Data
16. H2
O.ai
Machine Intelligence
1. The algorithm understands malicious behaviour through data
2. Algorithm is smart to work without features - metadata
3. Does not need alerts for training
4. Helps in identifying any kind of anomalous behaviour
5. Deeper insights about customer