This is the project to generate some Swiss Roll datasets.
When I wanted to search for some swiss-roll data sets with labels for machine learning on the Internet, I found that there were relatively few resources in this area, and it was difficult to find the appropriate source code to generate the corresponding data sets. Thus, I write the MatLab code to generate some kinds of swiss-roll. I hope this can help you.
The Swiss Rolls dataset is, itself, not very interesting. It is simply a standard "hello world" 3 dimensional data set which is universally known among academics working on various topics of dimensionality reduction techniques and algorithms.
The following is the classic swiss roll data set (OriginClassicData.m).
In my project, I focus on both 2 and 4 classes labeled data sets. I generate two different kinds of swiss roll data set. The one is derived from the classical swiss roll data set in which the data is generated uniformly (ClassicSwissRoll.m).
The another one is the broken swiss roll. I constructed the broken swiss data sets in python code (BrokenSwissRoll.py) and plot it with MatLab (pythondata_plot.m).
