「model-building」の検索結果

AI systematic knowledge acquisition from unfamiliar territory — deliberate model-building with feedback loops. Maps spaced repetition principles to AI reasoning: survey the territory, hypothesize structure, explore with probes, integrate findings, verify understanding, and consolidate for future retrieval. Use when encountering an unfamiliar codebase or domain, when a user asks about a topic requiring genuine investigation rather than recall, when multiple conflicting sources require building a coherent model, or when preparing to teach a topic and deep understanding is required first.

日本語の概要は準備中です。原文の説明を表示しています。

Designs discovery, modeling, and validation workflows for prognostic biomarkers in biomedical and clinical research. Always use this skill when the user needs a prognostic biomarker study blueprint rather than a diagnostic test protocol, predictive biomarker design, treatment recommendation, or a completed manuscript. Focus on endpoint family, follow-up horizon, time scale, candidate marker strategy, model-building logic, risk stratification framework, and internal/external validation requirements. Do not invent cohort size, event rate, assay readiness, literature support, or validation access.

日本語の概要は準備中です。原文の説明を表示しています。

Use this skill when the user wants to curate the SATK library knowledge index for a Simulink project — mark commonly used blocks, correct auto-assigned block categories, or improve block descriptions so the model-building agent picks better blocks from custom libraries. Triggered by phrases like "mark blocks common", "improve block descriptions", "correct block category", "make the agent prefer certain blocks", and from Gate 3 of building-simulink-models.

日本語の概要は準備中です。原文の説明を表示しています。