Timeline for Would PCA work for boolean (binary) data types?
Current License: CC BY-SA 3.0
21 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 8, 2021 at 16:20 | history | protected | Sycorax♦ | ||
| Jan 8, 2021 at 15:35 | comment | added | Ben Allen | Here is a 2020 reference for an R implementation of PCA on binary data: github.com/andland/logisticPCA arxiv.org/abs/1510.06112 | |
| Sep 28, 2020 at 0:05 | answer | added | Alf Pascu | timeline score: 11 | |
| Jul 21, 2019 at 12:11 | history | edited | kjetil b halvorsen♦ |
edited tags
|
|
| Jul 10, 2015 at 16:29 | vote | accept | Alvin Nunez | ||
| Jul 10, 2015 at 16:29 | vote | accept | Alvin Nunez | ||
| Jul 10, 2015 at 16:29 | |||||
| Jul 6, 2015 at 16:56 | vote | accept | Alvin Nunez | ||
| Jul 10, 2015 at 16:29 | |||||
| Jul 4, 2015 at 1:11 | history | tweeted | twitter.com/#!/StackStats/status/617138562434211840 | ||
| Jul 3, 2015 at 10:13 | comment | added | ttnphns | @Antoine, I definitely concur with your last part, but no so readily with the first. If I were to take strick stance I'd say it is factor analysis not PCA which could be seen as a way to cluster variables See, e.g. bullet 8 at the bottom here. PCA does not explain specifically associations between variables. | |
| Jul 3, 2015 at 9:54 | comment | added | Antoine | @ttnphns PCA can be viewed as a way to cluster variables. Also, PCA and cluster analysis can be used in sequence | |
| Jul 3, 2015 at 9:41 | answer | added | Antoine | timeline score: 18 | |
| Jul 3, 2015 at 7:11 | comment | added | ttnphns | Short answer: linear PCA (if it is taken as dimensionality reduction technique and not latent variable technique as factor analysis) can be used for scale (metrical) or binary data. Plain (linear) PCA should not be used, however, with ordinal data or nominal data - unless these data are turned into metrical or binary (e.g. dummy) some way. | |
| Jul 3, 2015 at 6:59 | comment | added | ttnphns |
a means of finding the similarity between individuals. But this task is for a Cluster analysis, not PCA.
|
|
| Jul 3, 2015 at 6:58 | comment | added | ttnphns | PCA on binary data such as yours ("present" vs "absent") would normally be performed without centering the variables because there is no reason to suggest the origin (the reference point) other than the original 0. So, instead of covariance- or correlation-based PCA we arrive at SSCP- or cosine-based one. Such analysis is very similar, almost equivalent to Multiple Correspondence analysis (= Homogeneity analysis) which could be the choice for you. | |
| Jul 3, 2015 at 6:50 | comment | added | ttnphns | This question is (almost) a duplicate of that one. PCA may be done on binary/boolean data, but doing factor analysis (including PCA "as if" it is FA) on such data is problematic. | |
| Jul 3, 2015 at 6:40 | history | edited | ttnphns | CC BY-SA 3.0 |
added 14 characters in body; edited tags; edited title
|
| Jul 3, 2015 at 4:43 | answer | added | Flounderer | timeline score: 13 | |
| Jul 2, 2015 at 21:30 | history | edited | Alvin Nunez | CC BY-SA 3.0 |
deleted 6 characters in body
|
| Jul 2, 2015 at 21:27 | answer | added | Vladislavs Dovgalecs | timeline score: 19 | |
| Jul 2, 2015 at 21:22 | review | First posts | |||
| Jul 2, 2015 at 21:24 | |||||
| Jul 2, 2015 at 21:20 | history | asked | Alvin Nunez | CC BY-SA 3.0 |