Exercise

Binarized scores

It's worthwhile to compare the performance of more than one anomaly detection algorithm before deciding which to use.

In this exercise, you'll construct a pair of binary anomaly scores based on local outlier factor (LOF) and the isolation forest. The isolation score vector iso_score generated in the previous exercise is preloaded for you to use.

Instructions 1/3

XP

1
2
3

Use scale() to scale the measurement columns of thyroid, assigning to scaled_thyroid_measurements.
Construct a LOF score for these measurements, setting k to 10 and saving the result as lof_score.

Take Hint (-10 XP)