[Corpora-List] repetitions

From: Alexander Osherenko (osherenko@gmx.de)
Date: Wed Nov 29 2006 - 11:36:43 MET

Next message: Djoerd Hiemstra: "[Corpora-List] SIGIR 2007: Call for Tutorials and Call for Workshops"

Previous message: Eckhard Bick: "Re: [Corpora-List] Spanish corpora and pos-taggers"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Hello!

I try to find out how I can create my dataset with possibly better performance and what attributes I use. I've created a dataset containing randomly chosen attributes that can repeat. Attributes with the same name have identical values. Hence, there is e.g. an attribute that is repeated 3 times.

I took the dataset with the highest absolute number of correct identified instances (identified with SMO), deleted all attribute repetitions and found out that the result is not identical with that before the deletion.

What could be the reason?

Best

Alexander Osherenko

-- 
"Ein Herz für Kinder" - Ihre Spende hilft! Aktion: www.deutschlandsegelt.de
Unser Dankeschön: Ihr Name auf dem Segel der 1. deutschen America's Cup-Yacht!

Next message: Djoerd Hiemstra: "[Corpora-List] SIGIR 2007: Call for Tutorials and Call for Workshops"
Previous message: Eckhard Bick: "Re: [Corpora-List] Spanish corpora and pos-taggers"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

This archive was generated by hypermail 2b29 : Wed Nov 29 2006 - 11:34:29 MET