Detecting fake accounts on a social-commerce platform

How would you identify fake accounts in our user base? What features and modeling approach would you use to build a classifier? How would you evaluate the model’s performance and monitor drift? What is the impact of fake users on the overall ecosystem (buyers, sellers, ads)? We found that 3% of accounts are fake—what conclusions or next steps does this number suggest?

Discuss feature engineering, supervised vs. unsupervised methods, precision-recall trade-offs, A/B testing for business impact.

Meta Data Scientist interview question: Identify Fake Accounts Using Machine Learning Techniques. {"blocks": [{"key": "7cbbddc9", "text": "Scenario", "type": "header-two", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {...