Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Analytics

ca Securecenter
Ten Commandments for Tackling Fraud:
The Role of Big Data and Predictive Analytics
Revathi Subramanian
SCX11S #CAWorld
SVP, Data Science
CA Technologies

2
© 2014 CA. ALL RIGHTS RESERVED.
Abstract
RevathiSubramanian
CA TechnologiesSVP, Data Science
Author of “Bank Fraud: Using Technology to Combat Losses”
Accurate enterprise-wide data combined with data-driven fraud analytics can have a transformational effect on banking and related industries. Revathiwill provide tips and insights on using technologies like neural network predictive modeling, user behavior-based pattern recognition and statistical big data analytics to reduce the risk of fraudulent activities in the enterprise.

3
Current Challenges with Fraud Management
FraudPrevention
Operational Costs
Customer
Experience

4
It’s a balancing act.
Achieving balance is the only way to increase customer acceptance.
Failing to achieve balance results in lost business and unhappy customers.
CUSTOMER
EXPERIENCE
FRAUD PREVENTION
OPERATIONAL COSTS
✔

5
1. Data: Garbage In; Garbage Out
“On two occasions I have been asked, ‘pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?’ … I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question.” Charles Babagge, ‘Passages from the Life of a Philosopher’

6
2. No documentation? No change!
“When dealing with data, a rose by a different name is not the same as a rose.” RevathiSubramanian, Bank Fraud: Using Technology to Combat Losses

7
3. Key employees are not a substitute for good documentation.
“Oh yeah! We made that change sometime in April and unfortunately we don’t have too many more details.”

8
4. Rules: More doesn’t mean better.
“As it is with children or pets, having too many rules is generally counterproductive. Too many rules tend to confuse them. Interestingly, fraud management systems are no different.” RevathiSubramanian, Bank Fraud: Using Technology to Combat Losses

9
5. Score: Never rest on your laurels.
“As fraud management systems get sophisticated, fraudsters also get sophisticated … Scoring processes have to keep on improving in order to tackle fraud effectively.” RevathiSubramanian, Bank Fraud: Using Technology to Combat Losses

10
Non-linear models are a requirement for any complex risk management problem.
Modeling technology used needs to look beyond the “Brownian motion.”
Needs to take into account the entire domain’s fraud patterns
Needs to constantly improve –automated yearly refreshes with regional tweaks
Advanced Analytical Scoring = Huge Value

11
6. Score + Rules = Winning Strategy
“A sophisticated scoring system combined with a limited set of rules to take into account operational considerations is the winning combination … Having too many rules can water down the fraud management system, but even the best scores are not very useful unless the operational use of the scores is streamlined with some carefully crafted rules… .” RevathiSubramanian, Bank Fraud: Using Technology to Combat Losses

12
Rules should be used to satisfy its business needs using a highly predictive score.
Rules can also implement business policy.
Business needs change over time and the rules need to change with that.
One size does not fit all; There should be no need to go back to the tailor to get the best fit either. 
Scores tell you who might not be legitimate.
Rules are what you decide to do with that knowledge.
Any system that requires you to outsource rules is a problem.
Maximum Precision with Maximum Flexibility

13
7. Fraud: It is everyone’s problem.
“Every little bit of information we drop on the floor, every transaction that doesn’t get recorded, every rule that doesn’t get used right, every score that doesn’t get used optimally, every fraud analyst that doesn’t get trained well has an impact on the overall fraud management picture.” RevathiSubramanian, Bank Fraud: Using Technology to Combat Losses

14
Online fraud problem is huge.
Digital footprint left behind by the customer is a goldmine of information that can be tracked and utilized with advanced analytics.
Device/browser characteristics can identify issues very quickly and precisely using advanced analytics.
Possible to track a device in real-time and nab the fraudster before major damage is done
Digital World: Goldmine of Information

15
9. Fraud Control Systems: If they rest, they rust.
“Fraud control systems have to evolve and change with the times, and the changes have to happen in an agile fashion.” RevathiSubramanian, Bank Fraud: Using Technology to Combat Losses

16
10. Continual Improvement: The cycle never ends.
“Every time there is a leap forward in the digital world, there is an equal leap forward in what the fraudsters can do to increase the losses.” RevathiSubramanian, Bank Fraud: Using Technology to Combat Losses

17
Consider the important factors in a payment transaction:
Each has unique history, distilled down to the critical information and combined with the transaction through a complex non-linear model to produce a score.
Beyond the Simple: Pivots Aperson, thing or factor having a major or central role, function or effect
Customer
Card
Device
Merchant
Customer
Card
Device
Merchant
Transaction
789

18
For More Information
To learn more about Security, please visit:
http://bit.ly/10WHYDm
Insert appropriate screenshot and textoverlayfrom following“More Info Graphics” slide here; ensure it links to correct page
Security

19
For Informational Purposes Only
© 2014CA. All rights reserved. All trademarks referenced herein belong to their respective companies.
This presentation provided at CA World 2014 is intended for information purposes only and does not form any type of warranty. Some of the specific slides with customer references relate to customer's specific use and experience of CA products and solutionssoactual results may vary.
Terms of this Presentation

Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Analytics

More Related Content

What's hot (16)

Viewers also liked (20)

Similar to Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Analytics (20)

More from CA Technologies (20)

Recently uploaded (20)

Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Analytics