Module 7:Data Representation
  Lecture 36: V-optimal Histograms
 

                                            

 

 

V-optimal histograms
  • A histogram of data with n bins
  • Reduce n to b where
  • Formally, assume a set V of n (sorted) values having frequencies respectively
  • Problem is to output another histogram H having b bins, i.e., b non-overlapping intervals on V
  • Interval is of the form and has a value
  • If value , estimate of is
  • Error in estimation is distance