Christine P Chai's contribution to the Discussion of ‘Vintage Factor Analysis with Varimax Performs Statistical Inference’ by Rohe & Zeng
Christine P Chai 对《Varimax 旋转的经典因子分析可进行统计推断》的讨论AI 解读 ↓
← 返回论文列表与筛选
阅读说明
本页整理公开书目信息，不生成学术质量评分。AI 解读根据题名及所注明的摘要或原文片段生成，未完成全文核验，可能有误；具体假设、证明与结论请核对原文。摘要保留来源文本，未经模型改写。
DOI 首次登记不等于接收或发表日期。部分更正、评论等条目没有独立摘要；缺少可靠材料时保留待补状态。
引用量采用 OpenAlex 单条关联记录，不合计多个记录；关联记录可能包含正式发表版本，不等于此版本单独引用。缺失不视为零，引用指标不代表论文质量。
关注作者按书目中的完整姓名匹配，同名和姓名变体需人工辨别。相关论文先按共享细分方向和具体方法排列，再参考通用方法与发表日期。
分类依据与纠错
AI 根据所提供的论文材料归类：材料讨论通用因子分析与稀疏主成分，未明确涉及宏观因子或金融因子模型。
反馈分类问题
正在核对关联版本…
代码、数据与补充材料
正在核对对应版本的资源…
关注这篇论文的作者
Christine P Chai 对《Varimax 旋转的经典因子分析可进行统计推断》的讨论
- 研究问题
- 这篇讨论关注因子分析的历史背景、书目语料的数据清洗、主题词解释、稀疏计算，以及稀疏主成分方法适用哪些应用。
- 方法与思路
- 讨论作者围绕被评论论文提出问题，结合主题词重复、词干处理和小样本高维应用例子，询问实现细节与方法选择依据。
- 材料陈述的贡献
- 所给材料的贡献是指出数据处理、因子解释和计算实现中需要澄清的问题；没有给出新的估计方法、定理或比较结果。
- 阅读关注点
- 应将讨论、主论文及作者回复分别阅读，核对语料预处理、稀疏矩阵实现，并查找对小样本高维适用性问题的实际回答。
- 阅读边界
- 这是讨论文章，非被评论主论文；OpenAlex 索引版本未注明，所提问题不能当作已证实缺陷或稀疏方法优于传统方法的证据。
原文摘要
Disclaimer: The opinions and views expressed here are those of the author and do not necessarily state or reflect those of Microsoft. Rohe and Zeng (2020) made an advancement to show that Vintage Factor Analysis with Varimax can perform statistical inference, with a wide range of applications. I appreciate that the authors made the whole vsp (Vintage Sparse PCA) package1 available in R for the general public. I have several questions regarding the paper: Why was Multiple Factor Analysis first developed by psychologists (Thurstone, 1935), instead of statisticians or computer scientists? Was there any historical context at that time? Could the authors briefly describe the data cleaning process of the academic bibliometrics corpus? In the topic modelling results, the word ‘biology’ appears in both factors 2 and 7, which is a bit confusing. Some factors also include nontechnical words like ‘on’ and ‘conference’. Finally, words of the same stem appear as separate words, such as ‘chemistry’ vs. ‘chemical’ and ‘economics’ vs. ‘economic’. The authors specified the compute resources for the code running time, indicating that Vintage Factor Analysis with Varimax can be implemented on a personal laptop within a reasonable time. Since the computation involves sparse matrices, are data structures like scipy.sparse.csr_matrix2 used in the code to save memory space? Are there particular types of applications that can benefit more from Vintage Sparse PCA, rather than just traditional PCA for dimensionality reduction? For instance, I investigated the relationship between the microbiome and environmental characteristics as a team (Beckman et al., 2015), and PCA projected the high-dimensional data to the first two principle components. Given the small n large p data, would Vintage Sparse PCA be more helpful than PCA?