Hi Zindi,

If the data is based on real occurrences, how is it possible that we can know if a policy that hasn't even started has lapsed. There's data for May (because there is a one month grace period), July and August in the test set. How is this being checked against prediction since the policy hasn't even taken effect?