I'm confused about the description of the inputs to the pca function between what the variables vs the observations are.
I have 4 vectors, a b c and d. Vectors a b and c contain values for corresponding variables a b and c. Vector d contains values for observations. That is, when a was a(1), then b was b(1), c was c(1), and the observed value was d(1). When a was a(2) then b was b(2), c was c(2), and the observed value was d(2). Etc etc. How do I structure this for the pca function input?

回答 (1 件)

Sagar
Sagar 2015 年 8 月 9 日

0 投票

Depending on your objective, you can probably consider 'd' as another variable. In such a case, put a, b, c, and d in four columns and conduct PCA. But if d is a dependent variable and a, b, and c are independent variables, you should probably conduct PCA on a, b, and c only by putting the data in three columns. So it depends on what you really want to to.

カテゴリ

ヘルプ センター および File Exchange で Dimensionality Reduction and Feature Extraction についてさらに検索

タグ

質問済み:

CP
2014 年 2 月 8 日

回答済み:

2015 年 8 月 9 日

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by