R help - May 2010 - bag.fraction in gbm package

Hi, Dear Greg,

Sorry to bother you again.

I have several questions about the 'gbm' package.

if the train.fraction is less than 1 (ie. 0.5) , then the* first* 50% will
be used to fit the model, the other 50% can be used to estimate the
performance.

if bag.fraction is 0.5, then gbm use the* random* 50% of the data to fit the
model, and the other 50% data is used to estimate the predictive
performance.

Is my understanding for train.fraction and bag.fraction right?  if not,what
is the difference?

can I set both fraction=1, and only use the cross.validation to select the
iterations?

Thanks so much!

-- 
Sincerely,
Changbin
--

	[[alternative HTML version deleted]]

Thanks, it really helps!


On Sat, May 1, 2010 at 2:34 PM, Ridgeway, Greg <gregr@rand.org> wrote:
> See friedman's paper "stochastic gradient boosting"
>
> Greg
>
> ------------------------------
>  *From*: Changbin Du
> *To*: Ridgeway, Greg
> *Cc*: r-help@r-project.org
> *Sent*: Sat May 01 14:20:23 2010
> *Subject*: bag.fraction in gbm package
> Hi, Dear Greg,
>
> Sorry to bother you again.
>
> I have several questions about the 'gbm' package.
>
> if the train.fraction is less than 1 (ie. 0.5) , then the* first* 50% will
> be used to fit the model, the other 50% can be used to estimate the
> performance.
>
> if bag.fraction is 0.5, then gbm use the* random* 50% of the data to fit
> the model, and the other 50% data is used to estimate the predictive
> performance.
>
> Is my understanding for train.fraction and bag.fraction right?  if not,what
> is the difference?
>
> can I set both fraction=1, and only use the cross.validation to select the
> iterations?
>
> Thanks so much!
>
> --
> Sincerely,
> Changbin
> --
>
>
>
> __________________________________________________________________________
>
> This email message is for the sole use of the intended...{{dropped:24}}