W. Yang
@wenjiany.bsky.social
Might be good to have a followup survey without paywall...
January 15, 2025 at 12:52 AM
Might be good to have a followup survey without paywall...
Nice example. Seems disjointed clusters in umap suggest they "can" be somehow distinguished, but doesn't indicate how much different they are... Limiting PCA to first two PCs is limited in this aspect.
December 24, 2024 at 1:14 PM
Nice example. Seems disjointed clusters in umap suggest they "can" be somehow distinguished, but doesn't indicate how much different they are... Limiting PCA to first two PCs is limited in this aspect.
Can we just say the probability that the true parameter overlaps with [L, U] is CI%? So we don't have to put either in the driver seat. It seems that both are aiming at Pr(overlap | (frequentist|bayesian) procedure). Pr(overlap | f) = Pr(overlap | f, data) for both frameworks?
December 7, 2024 at 4:33 PM
Can we just say the probability that the true parameter overlaps with [L, U] is CI%? So we don't have to put either in the driver seat. It seems that both are aiming at Pr(overlap | (frequentist|bayesian) procedure). Pr(overlap | f) = Pr(overlap | f, data) for both frameworks?
Is nature genetics, etc on body?
December 4, 2024 at 5:25 PM
Is nature genetics, etc on body?