In my last blog post, I analyzed the differences of metric vs. nonmetric MDS when applied to the NOUN data base. Today, I want to continue with showing some machine learning results, updating the ones from our 2018 AIC paper (see these two blog posts: part 1 and part 2).
Some time ago, I wrote two blog posts about a hybrid way for obtaining the dimensions of a conceptual space (see here and here). Currently, I’m rerunning these experiments in a more detailed way and today I want to share both the motivation for doing this as well as some first results.
Last time, I have shared the first results obtained by the LTN on the conceptual space of movies. Today, I want to give you a quick update on the first membership function variant that I have investigated.
After having already written a lot about Logic Tensor Networks, today I will finally share some first results of how they perform in a multi-label classification task on the conceptual space of movies.
I’ve already talked about InfoGAN  a couple of times (here, here, and here). InfoGAN is a specific neural network architecture that claims to extract interpretable and semantically meaningful dimensions from unlabeled data sets – exactly what we need in order to automatically extract a conceptual space from data.
InfoGAN is however not the only architecture that makes this claim. Today, I will talk about the β-variational autoencoder (β-VAE)  which uses a different approach for reaching the same goal.