第 10 / 11 页
非代码面试题
显示 20 / 215 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
2602Why Early Stopping Matters Even if Train Loss Falls 12Why can validation performance start to deteriorate even while the training objective of boosting keeps improving?机器学习中等essay未尝试免费2620A Bound on Total Function Movement 8Suppose every boosting round changes any one point's prediction by at most eta A in absolute value. What upper bound does this imply on the total prediction movement after M rounds?机器学习困难derivation未尝试面试订阅2645Why Global-Norm Clipping Preserves Direction 14Why does global-norm clipping change the magnitude of a gradient vector but not its direction whenever clipping is active?机器学习困难derivation未尝试面试订阅2649How Many Expanding-Window Folds Fit in a Monthly Panel?You have 60 months of data. Each expanding-window fold uses 24 months for training, the next 6 months for validation, and then advances by 6 months. How many validation folds fit?机器学习中等数值题未尝试免费2653Training Observations Per Fold in Grouped Cross-ValidationThere are 12 issuers and each issuer contributes 5 observations. In 3-fold grouped cross-validation, one fold holds out 4 issuers at a time. How many observations are used for training in each fold?机器学习中等数值题未尝试面试订阅2655Why Expanding Windows Can Beat Rolling Windows Under Sparse DataWhy might an expanding-window CV design be preferable to a rolling-window design when the series is short and drift is present but not violent?机器学习困难essay未尝试面试订阅2660Why Rare-Event Stratification MattersWhy can ordinary random folds become misleading in a rare-event problem even when the data are IID?机器学习困难essay未尝试面试订阅2662Why Overlapping Validation Windows Complicate Score AggregationWhy should a practitioner be careful when averaging performance over overlapping validation windows?机器学习中等essay未尝试面试订阅2665Why Tiny Folds Can Exaggerate RegularizationWhy can a very small training fold make heavily regularized models look better than they would on the full training set?机器学习困难essay未尝试面试订阅2670Why the Best CV Design Depends on the Deployment UnitWhy should the fold rule mirror the unit on which the model will actually generalize in production?机器学习困难essay未尝试面试订阅2674Precision of a Rare Alpha Event DetectorOnly 2% of days contain a true dislocation worth trading. A classifier catches 65% of those days but fires falsely on 4% of normal days. What is the precision of a positive alert?机器学习中等derivation未尝试面试订阅2675Break-Even Hit Rate After Trading CostsA directional model earns +1 unit on a correct trade and -1 unit on an incorrect trade before costs. Each round trip also pays a cost of 0.08 units regardless of outcome. What hit rate p makes expected net PnL zero?机器学习困难derivation未尝试面试订阅2676A Tiny Tail Probability Can Dominate Average PnLA strategy makes +0.04% on 98% of days and loses -2.5% on the remaining 2% of days. What is the unconditional average daily return?机器学习简单数值题未尝试免费2677Half-Life Discounting of Old Regime DataSuppose predictive relevance decays with a half-life of 6 months. Relative to current-regime data, what weight should you attach to observations that are 18 months old under this exponential-decay heuristic?机器学习中等derivation未尝试面试订阅2678Net Signal-to-Cost RatioA strategy's gross expected edge is 5 bps per trade, but execution-cost uncertainty has a standard deviation of 12 bps per trade. What is the gross-edge-to-cost-noise ratio?机器学习中等数值题未尝试面试订阅2684Why Short Windows Adapt but Also WhipsawWhy does a short rolling window often react faster to new regimes while simultaneously making parameter estimates much less stable?机器学习困难essay未尝试面试订阅2685Why Crisis Prediction Suffers From Tiny Relevant Sample SizeWhy does a century of daily data still leave very little effective evidence for training a model about true crisis behavior?机器学习困难essay未尝试面试订阅2689Why Adaptive Opponents Break Stationary Learning AssumptionsWhy is finance especially hostile to the idea that the data-generating process will sit still while your model learns it?机器学习困难essay未尝试面试订阅2690Why Fast Strategies Are More Sensitive to Small Modeling ErrorsWhy can a tiny forecast error or latency miss matter much more for a very fast strategy than for a slow rebalancing signal?机器学习困难essay未尝试面试订阅2695Why Two Similar-Looking Features Can Swap Usefulness Across RegimesWhy might one liquidity feature dominate in calm markets while another dominates in stressed markets, even though they looked redundant in a pooled sample?机器学习困难essay未尝试面试订阅