Our goal is to select variables in high dimensional settings. I need to separate variables into distinct groups, such that the within group correlation is less than some threshold. Can anyone help with some ideas other than k-means? PCA or ICA won't do as I need to select variables.
Our goal is to select variables in high dimensional settings. I need to separate variables into distinct groups, such that the within group correlation is less than some threshold. Can anyone help with some ideas other than k-means? PCA or ICA won't do as I need to select variables.
Thanks a lot!
请大家帮忙看看
请大家帮忙看看