第 3 / 5 页
非代码面试题
显示 20 / 82 道匹配题目
答题状态：未尝试未正确已正确
ID题目领域难度题型进度权限
4361Before Comparing CurvesBefore you compare ROC and PR curves across models, what dataset property should you check first?机器学习中等essay未尝试面试订阅4362Before RecalibratingA model looks poorly calibrated. What should you check first before concluding the model itself is broken?机器学习中等essay未尝试面试订阅4363Before Picking a ThresholdWhat is the first quantity you should pin down before optimizing a classification threshold?机器学习中等essay未尝试面试订阅4364Before Trusting AUCAUC improved a little after retraining. What should you ask first before declaring the new model practically better?机器学习中等essay未尝试面试订阅4365Calibration Is Not EnoughWhy should you hesitate before selecting a perfectly calibrated model that has weak ranking power?机器学习中等essay未尝试面试订阅4366Nested CV Fit Count 1Three model sizes have mean CV AUCs 0.790, 0.802, and 0.808. The standard error of the best score is 0.010. Under the one-standard-error rule, which is the simplest model you would keep?机器学习简单数值题未尝试面试订阅4367Nested CV Fit Count 2A search grid contains 4 learning rates, 3 tree depths, and 5 regularization strengths. How many hyperparameter combinations are in the grid?机器学习简单数值题未尝试面试订阅4368Nested CV Fit Count 3Successive halving starts with 27 configurations. Each round keeps one third of the configurations and evaluates all survivors once. If you run three rounds total, how many model fits are executed?机器学习简单数值题未尝试面试订阅4369Nested CV Fit Count 4You compare 12 hyperparameter settings with 5-fold CV repeated 3 times. How many validation scores are produced in total across all settings and folds?机器学习简单数值题未尝试面试订阅4370Nested CV Fit Count 5A time-series hyperparameter sweep evaluates 8 settings on 6 expanding-window splits. If each setting is refit once per split, how many model fits are needed?机器学习简单数值题未尝试面试订阅4371Successive Halving Budget 1A grid over regularization strength C expands from 5 log-spaced values to 9 log-spaced values, while all other settings stay fixed. If you use 4 values for another hyperparameter and 6-fold CV, how many additional model fits does the denser C grid create?机器学习中等数值题未尝试面试订阅4372Successive Halving Budget 2A random search draws 20 configurations independently, and the genuinely good region occupies 8% of the hyperparameter space. What is the probability of hitting that region at least once?机器学习中等数值题未尝试面试订阅4373Successive Halving Budget 3Successive halving starts with 64 configurations. Compare two keep ratios over three rounds total: keep one half each round versus keep one quarter each round. How many fewer fits does the one-quarter schedule use?机器学习中等数值题未尝试面试订阅4374Successive Halving Budget 4A tuning run tests 30 configurations. Switching from 10-fold CV to 5-fold CV while keeping the configuration set fixed saves how many model fits?机器学习中等数值题未尝试面试订阅4375Successive Halving Budget 5A random-search budget increases from 40 to 55 configurations. Each configuration uses 4-fold CV, and each fit takes 12 minutes. How much extra wall-clock training time is implied if runs are serial?机器学习中等数值题未尝试面试订阅4376Validation Set WorshipOnly 1% of cases are positive, and the desk can manually investigate just the top 100 alerts each day. When tuning thresholds, would PR-oriented metrics or ROC-oriented metrics deserve more emphasis first?机器学习中等essay未尝试面试订阅4377Grid Or Random SearchA feature standardization step is fitted on the full dataset before cross-validation instead of inside each fold. What is the main tuning problem with that workflow?机器学习中等essay未尝试面试订阅4378Early Stopping LeakageYou can afford only 30 evaluations, and experience suggests that only a few hyperparameters matter strongly while the rest are weak. Would grid search or random search usually deserve the first try?机器学习中等essay未尝试面试订阅4379Budget-Limited TuningData are scarce but the search space is broad, and you want an almost unbiased performance estimate after tuning. Is nested CV conceptually appropriate here despite its cost?机器学习中等essay未尝试面试订阅4380Fair Model ComparisonA researcher keeps rerunning the tuning loop until one hyperparameter setting looks best on CV by a tiny margin. What is the core risk in that behavior?机器学习中等essay未尝试面试订阅