Graduate Analytics Consultant at a tech services company with 51-200 employees
Real User
2020-11-25T23:22:20Z
Nov 25, 2020
The problem I was facing, from a machine learning perspective, it only had a supervised learning capability. You would have to provide your data live, but in fraud, the pattern of the fraudsters keeps changing and it's impossible to provide data labels. That's where the user unsupervised learning comes in handy — you don't have to tell them, "okay, this is fraud and this is not fraud." If unsupervised learning was also incorporated with Amazon SageMaker, that would be really cool. I am talking about anomaly detection algorithms, like isolation, forest, or anything on the neural network side for anomaly detection, including autoencoders. These are some things which companies would really like to use. There was also a problem with latency. In fraud detection, everything needs to be happening in real-time, but some of the algorithms ran for three to four minutes, which is not a viable option.
Find out what your peers are saying about Amazon Web Services (AWS), Riskified, Broadcom and others in Fraud Detection and Prevention. Updated: November 2024.
What is Fraud Detection and Prevention? It wasn’t that long ago that fraud detection and prevention involved reviewing a fair bit of historical data analysis. Data scientists would be poring over tons of credit card records in order to spot fraudulent (or with luck, potentially fraudulent) activity.
Fast forward to today and we see fraud detection systems depend on catching and stopping fraud the second it’s spotted or even before it actually occurs. Automated solutions for fraud...
The problem I was facing, from a machine learning perspective, it only had a supervised learning capability. You would have to provide your data live, but in fraud, the pattern of the fraudsters keeps changing and it's impossible to provide data labels. That's where the user unsupervised learning comes in handy — you don't have to tell them, "okay, this is fraud and this is not fraud." If unsupervised learning was also incorporated with Amazon SageMaker, that would be really cool. I am talking about anomaly detection algorithms, like isolation, forest, or anything on the neural network side for anomaly detection, including autoencoders. These are some things which companies would really like to use. There was also a problem with latency. In fraud detection, everything needs to be happening in real-time, but some of the algorithms ran for three to four minutes, which is not a viable option.