Latent space exploration using Generative Kernel PCA
Latent spaces can give us deeper insights into the structure of datasets and understand its features. In this talk we will investigate latent feature spaces extracted by kernel PCA by using a method known as generative kernel PCA. This generative mechanism will allow us to effectively explore and interpret this latent space by continuously generating new data points.
David Winant has recently completed his Master of Artificial Intelligence at the KU Leuven. Previously, he obtained a Master degree in Physics at the KU Leuven in 2018.