[prev in list] [next in list] [prev in thread] [next in thread]
List: wekalist
Subject: Re: [Wekalist] M5P's regression tree: information in the leaves
From: "Peter Reutemann" <fracpete () waikato ! ac ! nz>
Date: 2008-06-24 22:12:56
Message-ID: 548e07050806241512ia63ada7w6707aa19262e7706 () mail ! gmail ! com
[Download RAW message or body]
> I would need you help to figure out the meaning of the information present in the \
> leaves of the regression trees returned by the M5P algorithm. I would very much \
> appreciate any helpful hint.
> --------------------------
>
> What is the percentange value present in every leaf? How is this value defined?
See here:
https://list.scms.waikato.ac.nz/mailman/htdig/wekalist/2006-January/005878.html
> --------------------------
>
> Most of my leaves have values above 70 - 80 %. I assume, this does not represent a \
> great fit. Is my understanding correct? How can one improve this?
> --------------------------
>
> Also, I notice that, although I set the minimum number of instances per leaf, for \
> example, to 10, it might still happen that some leaves have less instances than my \
> threshold. Do you happen to know why?
See here:
https://list.scms.waikato.ac.nz/mailman/htdig/wekalist/2006-May/007159.html
Cheers, Peter
--
Peter Reutemann, Dept. of Computer Science, University of Waikato, NZ
http://www.cs.waikato.ac.nz/~fracpete/ Ph. +64 (7) 858-5174
_______________________________________________
Wekalist mailing list
Wekalist@list.scms.waikato.ac.nz
https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic