Reproducible experiments conducted in the paper 'Uncertainty Quantification in Anomaly Detection with Cross-Conformal p-Values'.
Conformal False-Discovery Control for Faithful Retrieval-Augmented Generation (ICML 2026)
Stochastic Gene Ontology Enrichment Analyses (GOEA) Simulations in manscript + Multiple-Test Correction Simulations
Generic enrichment analysis
Finding association between clinical, pathological and molecular features
Re-randomisation statistics toolkit in Python — Fisher's resampling test, pairwise multi-group comparisons with FDR / Bonferroni correction, binomial proportion tests with Wilson CIs, and a unified dispatcher for parametric / non-parametric hypothesis tests.
Multiple Hypotheses Testing for Discrete Data
Cuts false drift alarms from 35.9% to 4.1% of evaluation rounds while catching 75/75 injected drifts. A FastAPI sidecar running KS/chi-squared/PSI tests on live feature and prediction streams under a Benjamini-Hochberg alert budget, with a hysteresis state machine, Prometheus metrics and a Grafana dashboard.
19. Širi significance testing — van kontingencije (t/z, bootstrap testovi). Significance Testing - perm_fun / permutation diffs + p-value - t-test (Welch) - ANOVA / perm variance of group means - chi-square + resampling p-value. (perm, t, ANOVA, χ²). - χ² / z excess frekvencije vs uniform + Benjamini–Hochberg FDR.
5. Kontingencija / testovi veze χ², Fisher exact, G-test, Cramér’s V, phi, odds ratio, relative risk, McNemar, Cochran–Mantel–Haenszel, permutation test, bootstrap, jackknife, Benjamini–Hochberg / Bonferroni
To associate your repository with the benjamini-hochberg topic, visit your repo's landing page and select "manage topics."