The dataset is from the internal result of one of my projects, you can ignore the details of this, with X having dimension (6000, 100), y having dimension (6000, 10). 6000 the data sample number.
- Chen, Minmin, et al. "Marginalized denoising autoencoders for domain adaptation." arXiv preprint arXiv:1206.4683 (2012).