Skill UI

This comprehensive guide outlines best practices for creating key clinical display items, including Table 1 (baseline characteristics), Kaplan-Meier curves, Forest Plots, and CONSORT diagrams. It emphasizes statistical rigor (e.g., using standardized differences instead of P values for baseline), ensuring numbers-at-risk, and maintaining consistency across all reporting elements for high-impact medical publications.

A comprehensive guide for stress-testing the evidence supporting ML claims before submitting to top-tier academic conferences. Covers rigorous auditing of baselines, ablation studies, robustness checks, data leakage, and calibrating claims against evidence strength, ensuring scientific reproducibility.

A comprehensive guide for designing and reporting rigorous computational studies in Operations Research (OR) manuscripts. It covers best practices for selecting benchmark instances and strong baselines, applying statistical methods to stochastic output (e.g., confidence intervals), and adhering to mandatory code and data deposit standards (ORJournal workflow) to ensure full reproducibility.

A comprehensive guide for designing and auditing rigorous empirical evaluations for sensing systems, particularly Human Activity Recognition (HAR). It emphasizes best practices such as leave-one-subject-out cross-validation, using real human subjects under diverse conditions, reporting event-level metrics (e.g., Macro-F1) over raw accuracy, establishing fair baselines, and mitigating data leakage (e.g., subject or session leakage). This ensures that reported results are robust, generalizable, and defensible against academic skepticism.

This comprehensive guide details the rigorous methodology for designing, executing, and reporting experimental results in Information Retrieval (IR). It covers best practices for selecting appropriate test collections, choosing defensible metrics, performing proper statistical significance testing (e.g., paired t-tests), ensuring baseline fairness, and proactively addressing modern pitfalls like LLM contamination. Essential for publishing high-quality, auditable research in IR/NLP.

This guide provides a comprehensive framework for designing rigorous and defensible evaluation sections for machine learning and information retrieval papers, particularly for top-tier conferences like WSDM. It covers the four critical evidence quadrants—Effectiveness, Validity, Efficiency, and Robustness—and details necessary protocols, such as temporal splits, proper baseline selection (including heuristics), ablation studies, and hardware-aware performance reporting. Use this when auditing or writing the experimental section of a research paper.