sparse data

  1. E

    Probit regression with sparse data - Should i standardize data or ??

    I am trying to create a probit model (later other models like logit, cloglog and weirder stuff will be tried feel free to make suggestions!), the data is very sparse, ie. many of the coviarate are often 0 and they are often also of a binary structure (i have less than 15 explanatory variables)...
  2. A

    Multivariate regression with sparse data for a few independent variables

    In my analysis I have several independent variables. My dataset has 2000 records but for a few variables the data is sparse, i.e., information is available only for about 200 records. When I include these variables the regression runs only on 200 records and I am effectively losing information...