Tweak the cost function in such a way that the impact of false positives on cost

[ben11]

A company is hosting a free-to-play online game with over a million active users. The game profits by inducing players to spend money on buying loot boxes. A Machine Learning Specialist uses data from 500,000 random users to train an XGBoost model that predicts players who are likely to buy at least 5 boxes within a month based on age, gender, playing hours, engagement patterns, etc. The collected data contains 150,000 positive samples and 350,000 negative samples. The model has high accuracy on the training dataset but low on the test dataset.
Which methods could the Specialist do to rectify the problem? (Select TWO. & why?)

1. Increase the maximum depth of a tree.
2. Tweak the cost function in such a way that the impact of false negatives on cost value is higher than false positives.
3. Tweak the cost function in such a way that the impact of false positives on cost value is higher than false negatives.
4. Choose random samples of the training data and copy them to the test data.
5. Copy a subset of the positive samples and add noise to the copied data

I understand answer 5.

I don’t understand why the other correct answer is 3. How did we conclude Precision and hence FP is more expensive?
The company makes profit by selling loot boxes. FP is less expensive. FN is more expensive. isn’t it?

[Carlo-TutorialsDojo]

Hello ben,

Thanks for your feedback. I appreciate it.

The trade-off between FP and FN depends on the specific goal and constraints presented in the scenario (which I think is unclear right now). You’re right that if the goal is to maximize profits, reducing FN is more important, whereas if it’s more important to reduce costs associated with marketing to players who are not likely to buy, minimizing FP is more important.

We’ll tweak the scenario to provide more clarity and better understand the goals and constraints involved.

Let me know if this helps.

Regards,

Carlo @ Tutorials Dojo