Statistical theory, methodology, applications, machine learning, and computation. ← all categories
Re-analyze 100 published synthetic control studies from top economics journals. For each, systematically vary the donor pool: remove 1, 2, or 5 donors (all combinations up to 1000 draws).
Monte Carlo simulation (10,000 replications) of first-stage F-test, Cragg-Donald, and Kleibergen-Paap statistics for IV strength at N=50-5000. At N=200, the F>10 rule rejects a truly strong instrument (first-stage R²=0.
Microsimulation using Consumer Expenditure Survey (N=24,000 households) at carbon prices $25, $50, $100/tCO₂. At $50/tCO₂: urban burden 1.
Analyze 50,000 gig workers across 5 platforms (Uber, Lyft, DoorDash, Instacart, TaskRabbit) over 24 months. Monthly churn rate follows log-normal (μ=-2.
Collect delivery fee data from 3 platforms (DoorDash, Uber Eats, Grubhub) across 200 US cities over 6 months (2.4M transactions).
Analyze 12,000 workers across 84 firms using commute distance as instrument for remote work eligibility. OLS: remote workers 12.
Apply 5 TI methods (Monocle3, Slingshot, PAGA, Palantir, scVelo) to 3 gold-standard datasets with known ground truth (synthetic + lineage tracing). Pairwise Kendall τ between pseudotime orderings: mean 0.
Downsample 5 scRNA-seq datasets (10X Chromium) from 10,000 to 500 UMIs/cell. Cell cycle classification accuracy (Seurat, cyclone) degrades from 82% to 41%.
Compare 5 CUB metrics (CAI, tAI, ENC, CBI, RSCU) against protein abundance (PaxDb) in E. coli, S.
Quantify phylogenetic signal (Fritz-Purvis D statistic and Pagel's λ) across evolutionary rate classes in SARS-CoV-2, Influenza A/H3N2, and HIV-1. Signal decays exponentially with substitution rate: λ(r) = exp(-4.
Compare neutral drift model vs frequency-dependent selection for ARG frequency distributions in 3 databases (CARD, ResFinder, AMRFinderPlus) across 2,400 bacterial genomes. Neutral drift (Wright-Fisher with mutation) fits observed frequency spectra with KS p>0.
Compare CLR, ALR, ILR, and raw relative abundance on 4 published microbiome-disease association datasets (IBD, obesity, colorectal cancer, diabetes). The 'winning' method (highest number of significant associations at FDR<0.
Benchmark ML survival models (Cox-PH, RSF, DeepSurv, Cox-nnet) on genomics/transcriptomics/proteomics features vs TNM clinical staging alone across 12 TCGA cohorts (N=5,847). Mean C-index: clinical staging 0.
Apply rigorous statistical tests (Clauset-Shalizi-Newman framework) to degree distributions of 6 PPI databases (BioGRID, STRING, IntAct, MINT, DIP, HPRD). Power-law fits are rejected (p<0.
Batch effects are a major confounder in genomics, and multiple correction methods exist. We compare ComBat, limma removeBatchEffect, Harmony, scVI, and MNN on 5 paired RNA-seq datasets where the same biological comparison was performed in two independent batches.
Alternative polyadenylation (APA) has been proposed as a cancer biomarker, with studies reporting widespread 3'UTR shortening in tumors. We test whether APA changes are cancer-specific or tissue-specific by analyzing RNA-seq data from 8 TCGA cancer types across 5 tissue origins (4,200 tumor, 800 normal samples).
GC-content bias in microarray and RNA-seq platforms is well-documented but rarely corrected in differential expression analyses. We audit 20 widely-cited microarray datasets from GEO, applying a permutation-based test that evaluates whether the overlap between differentially expressed gene lists and GC-content-correlated genes exceeds chance.
Technical debt density distribution across projects is poorly understood. We analyze 8,247 projects (6 languages) via SonarQube.
Semantic segmentation quality measured by IoU treats all pixels equally, but boundary pixels are inherently ambiguous and annotator agreement drops to near-chance there. We propose Attention Map Entropy (AME) computed from self-attention maps at the penultimate layer of ViT-based segmentation models.
Vision Transformers were hypothesized to be more shape-biased than CNNs due to global attention, but findings are contradictory. We resolve this through Fourier-domain selective masking: removing spatial frequency bands from ImageNet images and measuring accuracy degradation.