Feature Learning for Nonlinear Dimensionality Reduction toward Maximal Extraction of Hidden Patterns

Demonstration Video, Supplemental Explanations and Evaluations, Source Code, Datasets

1. Demo Video

Video 1: A demonstration of the visual interface.


2. Supplemental Explanations and Evaluations

The supplementary document [pdf document].


3. Source Code

The source code of FEALM, the exmeplifying method, motivating examples, computational evaluations, and supplemental explanations [GitHub Link].

* The source code contains all detailed analysis processes related to the manuscript content.


4. Datasets

1. Two-spheres (Sect. 3: Motivating Examples): generate with the above source code or download from: [data].

2. 20 Newsgroups (Sect. 6: Computational Evaluation) [data source]. The dataset is also available in scikit-learn. To reproduce the computational evaluation results, generate preprocessed data with our sourcecode on GitHub.

3. Wine (Sect. 7: Case Study 1) [data source]. The dataset is also available in scikit-learn.

4. Cooperative Election Study Common Content, 2020 (Sect. 7: Case Study 2) [data source], [codebook], [processed data used for Case Study 2].

5. MNIST Handwritten Digits (Appendix E.4: Additional Case Study) [data source].

* If you use these datasets for your publication, please follow the citation policy of each data source.