Frank Abagnale Jr. First fraud before age of 16 Bank fraud, airline pilot, doctor and so on. 7-8 identities, fraud in 16 countries and got away from police custody twice, once from airplane. All this before age of 21. He was jailed for 5 years and helped FBI and then started his own company to help fraud prevention.
Slash n: Technical Session 7 - Fraudsters are smart, Frank is smarter - Vivek Mehta, Fareed Jawad
Fraudsters are smart, “Frank” is Smarter - Fareed and Vivek28/01/13 1
OutlineWhy detect fraud – Is there a problem?Why an intelligent system?How we built one 2 of 22
Show me some numbersWhat was the value of all electronic transactions globally for year 2012?$17 trillion (with a T)This includes all credit, debit and pre-paid cards used in both online and offline (card present) scenarios for purchases and cash withdrawals
More NumbersHow much of $17T was lost due to FRAUD?$8 billion in 2012, > $10 billion by 2015Fraud rate of 0.05% – Not too bad right?Wrong !!
Getting specificReminder - 0.05% ratio is for all transactions including face to face transactionsThe fraud rate is a much more scary 3.5% for Online transactions aka CNPGlobal e-Commerce is expected to exceed $1T in 2013 –> $3.5B will be lost due to fraudAdd to this, the erosion due to loss of future business from impacted customersBig Customer Impact ! Big deal for us !!
The Big Fight...Fraud to transaction ratio has been constant over the past 10 yearsThis ratio should not lull us into a false sense of security – bigger numbers are at stake and increasing as volumes growThe crooks LOVE e-Commerce (think 3.5%)How do we then figure out if a transaction is genuine or a victim of fraudIntelligently of course! - ENTER FRANK !!
Fraud Detection SystemTwo parts Signals/Features Algorithm 8 of 22
Rule based systemRules on various signals Num of transaction from a card in last one day Transaction amount and many moreThresholds are hand craftedFraud Score = sum of individual scores 9 of 22
Need for Smarter system Too much data for manual analysis Businesses are evolving Fraudsters are evolving Extending to really high dimension – pushing beyond limits of rule based system 10 of 22
Designing FrankLabeled data missingObservation Very few fraud records When you see one, you can identify oneSocial behavior 11 of 22
Visualization TotalAmount Number of transactions in a day by a user 12 of 22
Density based clusteringEpsilonMin-pts pStatistical distance p1 Scale invariant q Correlation taken into account 18 of 22
Clustering for detecting fraudCluster the data using density based clusteringFor new point find distance to all the existing clustersIf there exists min-pts with epsilon dist in a cluster, new point belongs to this clusterIf doesnt belong to any cluster -> fraud 19 of 22
Computing fraud probabilityWe find nearest clusterConvert the distance to probability using chi-square distributionProbability of fraud between 0 and 1 20 of 22
ExecutionDistributed clusteringReal-time model updating< 20ms to compute fraud probabilitySuspend the payment authorization in real time 21 of 22