{"blocks": [{"key": "2e17eb7a", "text": "Scenario", "type": "header-two", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "bf10f1b2", "text": "Phone-screen discussion with a hiring manager who wants to quickly verify a candidate’s breadth of machine-learning fundamentals for an e‑commerce recommendation platform.", "type": "unstyled", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "9ec4d29f", "text": "Question", "type": "header-two", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "107eedeb", "text": "Compare decision trees and random forests.", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "f2175c4a", "text": "Explain L1 vs L2 regularization and how each combats over- or under-fitting.", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "5e4e1901", "text": "With one million samples, would you choose DNN or KNN? Why?", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "d64578e0", "text": "Is the ROC curve defined only for binary classification? How would you plot one from a list of scores?", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "cfb88001", "text": "What causes training-loss oscillations and how would you address them?", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "6dbd97e7", "text": "Define data drift and describe how you would detect it in production.", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "125fed7f", "text": "Differentiate convex and non-convex objective functions.", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "65ae5bdd", "text": "Where do vanishing gradients typically occur in a neural network and how can you mitigate them?", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "bb4b1540", "text": "How does increasing decision-tree depth impact inference time (linear, logarithmic, exponential)?", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "9b5999a5", "text": "Cross-validation vs train_test_split – which is more robust and why?", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "b037a68c", "text": "Summarize the key ideas behind CNNs.", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "199de6aa", "text": "Contrast transformer encoders and decoders.", "type": "unordered-list-item", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "4f20380c", "t