Normally I try to toot apropos gopher writing or (non-)radio shows, but tbh right now
got clml.pca #pca working again (basically does an eigendecomposition using the BLAS but otherwise lisp-native)
Realised Eric Sandewall published a book on (#lisp) #AI approaches in 2014, and thinking about how it relates to their writing in 1981 and me.
=_=
I only have a key for my tilde.club gopher. I guess I could squirrel phosts there.
Mastodon, what is a good (freely available) dataset to give an example of principal component analysis?
Thanks for answering or sharing! #statistics#pca#rstats
Step 1: Load your data.
Step 2: Perform Principal Component Analysis (PCA).
Step 3: Calculate the variance explained.
Step 4: Create a stunning scree plot.
Step 5: Interpret the plot to find the "elbow."
Step 6: Decide how many components to retain.
Step 7: Apply your decision and get insights!
This news is particularly confusing perhaps to outsiders. You have to read into it and understand at least a little #church history. Most #presbyterians in Australia merged into the #UnitingChurch in 1977. What's left in the #Presbyterian Church in Australia is the 1/3 of the most "conservative." The American equivalent would be something like the #PCA announcing that they are banning land acknowledgments at services. Bottom line, why does a national church body 1/2
The basic rationale is to use random split-half data to identify what's "true" versus sampling error. Scores are based on similarities between eigenvectors or cluster centres, rather than, e.g., the shape of the eigenvalue plot.