3

I was studying classification algorithm , LDA (Linear Discriminant Analysis). In this, we have to estimate the mean and variance of predictors, given a particular class. Can I use Bayesian estimation procedure to estimate mean and variance from the data ? Just wanted to know, how much results vary when compared to when we use conventional estimates of mean and variance from the data and then using them for classification procedures.

I am a beginner in this field, so pardon my naive question.

Thanks

Gavin Simpson
  • 37,567
  • 5
  • 110
  • 153
lovekesh
  • 459
  • 5
  • 16
  • Of your possible interest, http://stats.stackexchange.com/q/31366/3277. – ttnphns Jan 06 '14 at 05:37
  • 1
    @ttnphns ironically, i was actually looking at this page moments before your reply. The link doesn't talk about use of bayesian estimation for mean and variance of class predictors. – lovekesh Jan 06 '14 at 05:57

0 Answers0