[Multivariate analysis]

1 次查看（过去 30 天）
Hotbirdym
Hotbirdym 2011-4-30
Hi,
I would like to ask you which method I should apply in order to establish a relation between a set of parameters and a certain measure which is computed regarding this set of parameters. My problem is the following: I have a set of 8 parameters which have an influence over a measure. For these parameters I have a set of observations. First I've tried to use PCA, but then I stopped because I thought it is not correct. I've computed the correlation matrix ( which is the same as the covariance matrix when the variables are standardised ) and I saw that I have no relation between those parameters ( the correlation is 0 ), but there is a negative correlation for each parameter and the measure that I compute. Thus, I've change PCA with another multivariate analysis method: Multiple Linear Regression (this method try to find how a set of independent parameters influence a dependent variable). But, when I've computed this method I saw that the weights assigned to each variable are very low ( such as 0.17, 0.10 and so on ) and I tend to believe that I do something wrong with my data.
I would like, if you don't mind, to give me a hint about which method I could use.
Thank you! Elena

请先登录，再回答此问题。

回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Dimensionality Reduction and Feature Extraction 的更多信息

标签

Community Treasure Hunt

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

Start Hunting!

Translated by