Free AWS MLA-C01 Actual Exam Questions - Question 1 Discussion

Probably C. Increasing target_precision is the only option directly tied to reducing false positives since precision measures how many predicted positives are actually correct. A and B don’t really impact false positive rates directly, just model complexity or training time. D makes no sense because switching to a regressor won’t help with classification errors like false positives. The question is about minimizing false positives, and focusing on precision is the most straightforward way.

C imo, since increasing target_precision directly reduces false positives by emphasizing precision.

Maybe C makes the most sense since target_precision is designed to focus on precision, which means fewer false positives. B and A don’t specifically address false positives, and D isn’t a classification approach.

Makes sense to avoid D since regression doesn't suit classification. I think increasing target_precision (C) directly targets false positives, unlike weight decay or epochs. So, C seems right.

I’d rule out D because regression won’t directly minimize false positives in classification. B might help with model learning but doesn’t specifically target false positives. C seems more focused on precision, so it fits better here.

C makes sense since higher target_precision prioritizes fewer false positives.

Option A isn’t great because zero weight decay can cause overfitting and may actually increase false positives. Better to control model complexity than risk that.

A/B? Setting weight decay to zero (A) could lead to overfitting, which might actually increase false positives by making the model too sensitive. Increasing training epochs (B) might help the model learn better patterns, but more epochs can also cause overfitting if not monitored, so it’s a bit risky. Changing to regressor (D) doesn’t really make sense here since it’s a classification problem. So between these, B seems like it could help indirectly by improving the model’s generalization if done carefully, but it’s not as directly targeted at false positives as adjusting precision in C.

Yeah, option C fits because increasing target_precision pushes the model to be more careful when labeling positives, which cuts down on false positives. So I’d go with C.

Makes sense to focus on precision when minimizing false positives since precision measures the accuracy of positive predictions. So C seems like the best bet here.

Wait, is there any info on how the model’s currently performing or what the distribution of classes looks like? Also, does the question say anything about which type of error (false positive vs false negative) is more costly? That might help figure out if adjusting target_precision (C) is enough or if another approach is better.