Homework 4

• Using the unpaired t-test to explain why in the following competition , the approach RCO10310310 is not significantly more accurate.

  Data are in the following zip file if you find it's difficult to connect to the above site. Thanks to Yien Ou for preparing it.

• Redo the same experiment of k-NN in the following paper:

  T. Joachims, Text Categorization with Support Vector Machines: Learning with Many Relevant Features. LS8-Report 23, Universitat Dortmund, LS VIII-Report, 1997.

  Note that in this paper a precision/recall-breakeven point is used.