Modern Scala Projects

Fraud detection approach

The following diagram illustrates the high-level architecture for our fraud detection pipeline:

High-level architecture for the fraud detection pipeline

The following is a quick overview of the fraud detection process:

1. First, we compute statistics on the training set, which serves as a cross-validation dataset. We are interested in the mean and standard deviation from the statistical numbers.
2. Next, we want to compute the net probability density function (PDF) for each sample in our cross-validation set.
3. We derive the net probability density as a product of the individual probability densities. 
4. Inside the algorithm, ...