Forum Home  >  perClass  >  Technical support  >  Thread
Search  
  
  Advanced Search
Welcome Guest   ·   Login   ·   Register   ·   Member List
Username Remember Me?
Password   forgot password?
   
   
Using sdalg in sdcrossval
 
Mathieu
Posted: 30 June 2010 09:07 PM   [ Ignore ]  
Newbie
Rank
Total Posts:  20
Joined  2010-02-18

Hi, I got the following message when I use my sdalg algorithm in sdcrossval:

??? Error using ==> sde_rotation at 185
Algorithm does not provide decision list object using tolist section

Error in ==> C:\Program Files\MATLAB\R2009b\toolbox\prsd_toolbox\sdcrossval.p>sdcrossval at 119

I know that the algorithm needs to return the list of all possible decisions but I don’t know how to do it. Could you give me an example returning such a list.

Thanks

Mathieu

Profile
 
 
pavel
Posted: 01 July 2010 08:52 AM   [ Ignore ]   [ # 1 ]  
Administrator
Avatar
RankRankRankRank
Total Posts:  240
Joined  2008-04-26

Dear Mathieu,

you may add the tolist section as another elseif part of your algorithm:

elseif tolist(alg)

     out=sdlist('apple','banana');

If your algorithm handles different problems, you may return the list of decisions corresponding to your classifier. If you use PRSD Studio operating points, it’s the pd.list where pd is your pipeline returning decisions or operating point.

Actually, if you can represent your algorithm in execution by a pipeline returning decisions, you only need to include toconvert section that returns this pipeline. sdcrossval will then convert your trained algorithm and take the list of possible decisions from the pipeline.

Does it help?

Pavel

Profile
 
 
   
 
 
‹‹ Classifier cascade without random data division      support vectors in sdsvc ››