search

epiverse-connect / epiverse-map

MIT License
1 stars 0 forks source link

Projection / dimension reduction technique #10

Open chartgerink opened 4 months ago

chartgerink commented 4 months ago

This issue is a follow up from our meeting on Thursday, August 1st.

We need to further discuss what projection/dimension reduction technique we implement. Mentioned during the meeting were Principal Component Analysis (PCA) and t-distributed stochastic neighbor embedding (tsne).

Bisaloo commented 3 months ago

The link is in the README but for completeness here, this blog post on a recipes map contains a comparison of PCA & UMAP: https://tomsing1.github.io/blog/posts/vectorsearch

paulkorir commented 2 months ago

Not sure this is the right place to post this but I'm wondering whether the visualisation should be static or dynamic with each search. An example of this is shown at https://www.sbert.net/examples/applications/semantic-search/README.html.

chartgerink commented 2 months ago

@paulkorir see also #6 😊