T-distributed Stochastic Neighbor Embedding (t-SNE) is a machine learning algorithm for visualization developed by Laurens van der Maaten and Geoffrey Hinton.
This would be a useful addition alongside PCA, and LDA.
It is a nonlinear dimensionality reduction technique well-suited for embedding high-dimensional data for visualization in a low-dimensional space of two or three dimensions.
Please can we have a node for this.
Thanks,
Simon.
Brilliant, I’ve now installed it, thanks for the pointer.
Simon.
Albeit this node only exposes a limited set of parameters compared to say some python implementations. Especially learning rate can’t be set. Hence I still rely on python based t-sne.
Hi @beginner et al -
We have a ticket open for t-SNE (L. Jonsson) node enhancements. I’ll add your comment on exposing additional parameters there as a +1. (Internal reference: AP-12930)
This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.