Statistics
62 problems
- Autocorrelation functionMedium
- Bayesian A/B Test (Beta-Binomial)Medium
- Bootstrap Confidence Interval (Percentile)Medium
- Bradley-Terry Strength MLEHard
- Chi-Square Test of IndependenceMedium
- Cohen's kappa scoreMedium
- Covariance matrixEasy
- CUPED Variance ReductionMedium
- DBSCAN clusteringMedium
- DDM Concept Drift DetectorMedium
- Delta Method for Ratio Metric VarianceHard
- Difference-in-DifferencesMedium
- Elo Rating UpdateEasy
- Fisher-Yates shuffleEasy
- Fleiss' KappaMedium
- Generate random subsets (bootstrap)Easy
- How big is the nucleus?Easy
- Hash Bucket A/B AssignmentEasy
- Holt-Winters Exponential SmoothingMedium
- Huber & Hinge lossesEasy
- Team-Draft Interleaving (Online Ranking Eval)Medium
- Inverse-transform samplingEasy
- Isolation Forest Anomaly ScoreMedium
- k-fold cross-validationEasy
- Kendall's TauMedium
- K-means: one iterationMedium
- Kolmogorov-Smirnov StatisticMedium
- LinUCB Contextual BanditHard
- Logistic regression — likelihood of the whole datasetEasy
- Logistic regression — probability of the true labelEasy
- Mann-Whitney U TestMedium
- Multiple-Testing Correction (Bonferroni / BH-FDR)Medium
- Mutual informationMedium
- Negative Sampling (Implicit Feedback)Medium
- pass@k Unbiased EstimatorMedium
- Pearson correlation coefficientEasy
- Pinball (Quantile) Regression LossMedium
- Point-in-Time Feature JoinMedium
- Pointwise mutual information (PMI)Medium
- Population Stability Index (PSI)Medium
- Rejection samplingMedium
- Repetition penalty (HuggingFace-style)Medium
- Reservoir samplingMedium
- Sample Ratio Mismatch (SRM) CheckMedium
- Sample Size for Statistical PowerMedium
- History beats a single point — next-step predictionEasy
- Build sliding-window training pairsMedium
- Sequential Test (mSPRT-Style Always-Valid p-Value)Hard
- Shannon entropyEasy
- SMOTE OversamplingMedium
- Spearman Rank CorrelationEasy
- Speculative decoding verificationMedium
- Temperature scalingEasy
- Temporal Train/Val SplitEasy
- Thompson Sampling (Bernoulli-Beta)Medium
- Time-series anomaly detectionMedium
- Top-k samplingEasy
- Top-p (nucleus) samplingMedium
- Two-Sample Test + p-valueMedium
- Wasserstein distance (1-D)Medium
- Weighted multinomial samplingEasy
- Wilson Confidence Interval (Proportion)Easy