Modern Scala Projects

Calculating the best error term and best F1 score

In this section, we will write a function to calculate:

• The best error term (also known as the Epsilon)
• The best F1 score

We start by defining a function called errorTermCalc. What parameters does it require? It is apparent that we need two parameters: 

• Our cross-validation dataset—DataFrame
• DataFrame, containing probability densities

There you go. We now have a function called errorTermCalc that takes two parameters and returns both the best error term and the best F1.

Why are these numbers important? To answer that question, we want to detect the outliers first. These are labeled data points that indicate fraud. Generating a new dataframe that classifies labeled data points as either ...