[Home]   [  News]   [  Events]   [  People]   [  Research]   [  Education]   [Visitor Info]   [UCSD Only]   [Admin]
Home > Events > CCoM > Abstract
Search this site:

Randolph E. Bank
Philip E. Gill
Michael Holst

Administrative Contact:
Terry Le

Office: AP&M 7431
Phone: (858)534-9813
Fax: (858)534-5273
E-mail: tele@ucsd.edu
The Disputed Federalist Papers: Resolution via Support Vector Machine Feature Selection

Olvi Mangasarian


In this talk we utilize a support vector machine feature selection procedure via concave minimization to solve the well-known Disputed Federalist Papers classification problem. First we find a separating plane that classifies correctly all the training set consisting of papers of known authorship, based on the relative frequencies of three words only. Then, using this three-dimensional separating plane, all of the 12 disputed papers ended up on one side of the separating plane. Our result coincides with previous statistical and combinatorial method results.

Tuesday, January 21, 2014
11:00AM AP&M 2402