第 26 / 32 页
非代码面试题
显示 20 / 622 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
4385Early-Stopping PatienceOne CV fold is much smaller than the others and dominates the variance of the average score. What tuning-related design concern should you address first?机器学习中等essay未尝试面试订阅4386Before TuningTraining AUC is very high but CV AUC is near chance. Before trying more hyperparameter values, what is the first diagnostic step?机器学习中等essay未尝试面试订阅4387Before Adding More DimensionsTwo hyperparameter settings differ in mean CV score by only 0.001, while the estimated standard error is 0.010. What is the first sensible interpretation?机器学习中等essay未尝试面试订阅4388Before Reporting Best ScoreA time-series tuning run says a very short lookback window wins, but recent live performance has deteriorated sharply. What should you inspect first before widening the search?机器学习中等essay未尝试面试订阅4389Before Choosing Nested CVA categorical encoder was fit once on all rows and then reused inside cross-validation. What is the immediate correction?机器学习中等essay未尝试面试订阅4390Before Blaming The SearchThe current best setting sits at extreme values on both the learning-rate and regularization grids. What should your next search action be?机器学习中等essay未尝试面试订阅4391Return Feature Alignment 1A stock closes at 100 yesterday, opens at 102 today, and closes at 101 today. What are the overnight return and the intraday return for today?机器学习简单数值题未尝试面试订阅4392Return Feature Alignment 2A stock returns 1.4% today while the benchmark returns 0.5%. If the stock's beta to the benchmark is 1.6, what market-adjusted residual return do you attribute to the stock?机器学习简单数值题未尝试面试订阅4393Return Feature Alignment 3At today's close, you build a leakage-safe trailing-mean return feature from the last five completed daily returns: [1%, -2%, 0%, 3%, 2%]. What feature value do you store?机器学习简单数值题未尝试面试订阅4394Return Feature Alignment 4A realized-volatility feature is defined as the root-mean-square of the last four daily returns. If those returns are [1%, -1%, 2%, 0%], what realized volatility feature do you get?机器学习简单数值题未尝试面试订阅4395Return Feature Alignment 5A momentum feature is defined as trailing 20-day cumulative return divided by trailing daily volatility. If cumulative return is 6% and daily volatility is 1.5%, what vol-scaled momentum value do you store?机器学习简单数值题未尝试面试订阅4396Cross-Sectional Z-Score 1If today's close is 100, tomorrow's open is 98, and tomorrow's close is 99, what is the next-day open-to-close return that would be used as an intraday label available after tomorrow's session?机器学习简单数值题未尝试面试订阅4397Cross-Sectional Z-Score 2An asset returns 1.2% today while the cross-sectional mean return of its universe is 0.4%. What demeaned return feature does the asset receive?机器学习简单数值题未尝试面试订阅4398Cross-Sectional Z-Score 3A stock returns 1.5% while its sector index returns 0.9%. If the stock's sector beta is 1.2, what sector-residual return feature do you compute?机器学习简单数值题未尝试面试订阅4399Cross-Sectional Z-Score 4Yesterday's return was 1.8%. The trailing mean of completed daily returns is 0.3% and the trailing standard deviation is 0.5%. What lagged return z-score feature do you record?机器学习简单数值题未尝试面试订阅4400Cross-Sectional Z-Score 5A stock closes at 50 yesterday and opens at 51 today, while the market index closes at 2000 yesterday and opens at 2020 today. If the stock's overnight beta to the market is 1.5, what market-adjusted overnight return feature do you compute?机器学习简单数值题未尝试面试订阅4401Close-to-Close LeakYou predict tomorrow's close-to-close return at today's close, but one feature uses today's official closing auction price finalized after the decision timestamp. Why is that a leakage problem?机器学习中等essay未尝试面试订阅4402Target Overlap TrapYou build a 5-day forward-return label every day and then use adjacent samples as if they were independent. What is the structural issue?机器学习中等essay未尝试面试订阅4403Rolling Mean That PeeksA feature at time t uses a rolling mean computed from t-19 through t+1. Why is that unacceptable even if it is only one extra day?机器学习中等essay未尝试面试订阅4404Cross-Sectional Residualization TimingWhy should factor residualization for a daily return feature use exposures known at the feature timestamp rather than exposures estimated using later returns?机器学习中等essay未尝试面试订阅