ahp-weighting
$
npx mdskill add yogsoth-ai/de-anthropocentric-research-engine/ahp-weighting使用 AHP 层次分析法确定评分维度权重,输出权重向量。
SKILL.md
.github/skills/ahp-weightingView on GitHub ↗
---
name: ahp-weighting
description: "SOP: 使用 AHP 层次分析法确定评分维度权重,输出权重向量"
version: 1.0.0
category: hypothesis-formation
type: sop
campaign: gap-prioritization
input: "维度列表(字符串数组)+ 可选的两两比较偏好矩阵"
output: "AHPWeights — 权重向量、一致性比率(CR)及判断矩阵"
dependencies:
skills:
- subagent-spawning
---
# AHP Weighting
使用 AHP 层次分析法确定评分维度权重,输出权重向量。
## HARD-GATE
<HARD-GATE>
- 输入维度数量必须在 [2, 9] 范围内(AHP 适用范围)
- 输出权重向量各元素之和必须等于 1.0(允许 ±0.001 误差)
- 一致性比率 CR 必须被计算并报告;若 CR > 0.1 必须标记警告
</HARD-GATE>
## Pipeline
1. **前置检查**: 验证维度列表非空且数量在 [2, 9] 范围内
2. **维度列表确认**: 输出维度列表供调用方确认;若已提供比较矩阵则跳至步骤 4
3. **两两比较矩阵构建**: 对每对维度 (i, j) 赋予 Saaty 标度值(1-9);矩阵满足 a[j][i] = 1/a[i][j]
4. **特征向量计算**: 对每列归一化后取行均值,得到优先级向量(权重)
5. **一致性比率检验**: 计算最大特征值 λ_max → 一致性指数 CI = (λ_max - n)/(n-1) → CR = CI/RI(查 Saaty RI 表);CR < 0.1 为可接受
6. **输出**: 返回 AHPWeights 对象;若 CR > 0.1 附加修正建议
## Output Format
```json
{
"dimensions": ["importance", "feasibility", "novelty", "impact"],
"comparison_matrix": [[1, 3, 2, 2], [0.33, 1, 0.5, 0.5], [0.5, 2, 1, 1], [0.5, 2, 1, 1]],
"weights": { "importance": 0.40, "feasibility": 0.15, "novelty": 0.23, "impact": 0.22 },
"lambda_max": 4.02,
"ci": 0.007,
"ri": 0.90,
"cr": 0.008,
"cr_acceptable": true,
"warnings": [],
"revision_suggestions": []
}
```
More from yogsoth-ai/de-anthropocentric-research-engine
- abductive-hypothesis-generationStrategy: 面对异常的最佳解释推理
- ablation-brainstormRemove components one by one, observe system changes to reveal hidden dependencies and generate ideas from structural gaps.
- ablation-component-mappingMap system architecture to ablatable units for ablation studies
- ablation-designDesign ablation studies to isolate component contributions in ML systems
- ablation-executionRemove components one by one from a system, record the response/impact of each removal.
- abp-vulnerability-classificationClassify assumptions on 2 axes — load-bearing (how much conclusion depends on it) × vulnerable (how likely to be false). Focuses attention on High-Load × High-Vulnerable quadrant.
- abstraction-extractionExtract abstract principles from concrete domain cases. Strips domain-specific details to reveal transferable mechanisms.
- abstraction-ladderPerform bisociation at multiple abstraction levels
- abstraction-ladderingMove between concrete and abstract framings — 3 levels up (Why?) and 3 levels down (How?) to find the most productive research level.
- abstraction-to-designAbstract biological principle to design principle. Bridge from biology to engineering.