Self-Organizing Maps and an Expertise Index for Transparent and Reproducible Consensus-Based Indicator Selection

This dataset contains supplementary tables used in the analysis of evaluator expertise and indicator clustering with Self-Organizing Maps (SOMs). Each table reports the hyperparameter settings and performance metrics for SOM training, including:

Initial and final sigma values
Initial and final learning rates
Initial, final, and minimum quantization error
Number of neurons used
Distribution of indicators across neurons

The dataset supports the reproducibility of the clustering analysis presented in the manuscript Self-Organizing Maps and Evaluator Expertise for Transparent and Reproducible Consensus-Based Indicator Selection.

Future versions of this dataset will include additional processed files and Jupyter notebooks used in the analysis.

本数据集包含用于借助自组织映射（Self-Organizing Maps, SOMs）开展评估者专长与指标聚类分析的辅助表格。每张表格均记录了自组织映射训练的超参数设置与性能指标，具体包括：初始与最终σ值、初始与最终学习率、初始、最终及最小量化误差、所用神经元数量、各指标在神经元间的分布情况。本数据集可复现发表于论文《自组织映射与评估者专长：实现透明且可复现的基于共识的指标选择》中的聚类分析研究。本数据集的后续版本将新增分析过程中使用的额外处理文件与Jupyter笔记本。