4 ms·

Recommendation: on your website [0] you mention a case study for detecting financial fraud. It seems to talk about what could be done rather than what has been

by trendia 10y ago

Recommendation: on your website [0] you mention a case study for detecting financial fraud. It seems to talk about what could be done rather than what has been done. So, I think you may benefit by showing specific examples of catching fraud. For instance: what is your performance compared to, say, human auditors? What sort of features can you find that human auditors cannot? And what is your false positive / true positive (ROC) like? [0] https://skymind.io/finance https://skymind.io/finance

vonnik 10y ago
Hi there! A lot of the fraud detection we do is actually anomaly detection, where we're not training a classifier on a labeled dataset, but instead creating a model that can surface unusual behavior. The subset of data that we surface is then passed to human analysts. What they care about, generally, is a low false positive rate, because false positives waste their time. We talk about this a bit more here: https://skymind.io/case-studies/orange https://skymind.io/case-studies/orange Sorry if that doesn't quite answer your question.
- yingxie3 10y ago
  For fraud detection such as what you did for orange, is there a rule of thumb on how much data one would need? Before heading down the path I would like to know whether my problem is solvable given the amount of data I have. Thanks in advance!
  - agibsonccc 10y ago
    We were in the terabytes. A viable proof of concept could be built with a few gigs of transactions to validate the idea though. The main bit with orange that isn't in the marketing material: We used unsupervised methods for this, not supervised.
agibsonccc 10y ago
Hi, I agree we could spend some time on numbers. We are actively working on this. The problem with enterprise customers like banks is the ability to put things in public. This includes things like accuracy..I agree we need a lot more of that in our material though. Thanks for the suggestion.

by trendia 10y ago

vonnik 10y ago

Hi there! A lot of the fraud detection we do is actually anomaly detection, where we're not training a classifier on a labeled dataset, but instead creating a model that can surface unusual behavior. The subset of data that we surface is then passed to human analysts. What they care about, generally, is a low false positive rate, because false positives waste their time. We talk about this a bit more here: https://skymind.io/case-studies/orange https://skymind.io/case-studies/orange Sorry if that doesn't quite answer your question.

yingxie3 10y ago
For fraud detection such as what you did for orange, is there a rule of thumb on how much data one would need? Before heading down the path I would like to know whether my problem is solvable given the amount of data I have. Thanks in advance!
- agibsonccc 10y ago
  We were in the terabytes. A viable proof of concept could be built with a few gigs of transactions to validate the idea though. The main bit with orange that isn't in the marketing material: We used unsupervised methods for this, not supervised.

agibsonccc 10y ago

Hi, I agree we could spend some time on numbers. We are actively working on this. The problem with enterprise customers like banks is the ability to put things in public. This includes things like accuracy..I agree we need a lot more of that in our material though. Thanks for the suggestion.