Also, feature_importances will tell you about how much the random forest is
relying on particular features. For instance, you could keep only half the
features - those that have the largest importances. You may have to set
compute_importances=True depending on which version you're running.
On Tue, Aug 13, 2013 at 4:00 AM, Gilles Louppe <[email protected]> wrote:
> Hi,
>
> As Roland says, this is a Numpy question rather than a scikit-learn
> question. If you want to ignore specific fields then it indeed amounts to
> removing the corresponding columns in your X array before feeding it to
> your estimator.
>
> (Note however that Random Forests have the advantages of being robust with
> respect to noise attributes. Training with or without shouldn't change the
> result by much.)
>
> Best,
>
> Gilles
>
>
> On 13 August 2013 12:24, Jason Williams <[email protected]> wrote:
>
>> Would it have influence on the result if some attributes are removed?
>> Just want to double check because I am not very familiar with machine
>> learning. If no impact for the result after some fields are taken out of
>> input training data, I will just remove those fields because that would be
>> simpler.
>>
>> Thank for help
>>
>>
>>
>>
>>
>> ________________________________
>> From: Roland Szabo <[email protected]>
>> To: Jason Williams <[email protected]>;
>> [email protected]
>> Sent: Tuesday, 13 August 2013, 6:11
>> Subject: Re: [Scikit-learn-general] Can Random Forest Classifer ignore
>> specific fields?
>>
>>
>>
>> Isn't it simpler to just take out the attributes you want to ignore from
>> the data?
>>
>> Assuming the numpy array a contains your data and you don't want the
>> first column (attribute):
>>
>> a = a[:,1:]
>>
>>
>>
>>
>> On Tue, Aug 13, 2013 at 1:03 PM, Jason Williams <
>> [email protected]> wrote:
>>
>> I follow an example found on the internet (
>> http://blog.yhathq.com/posts/random-forests-in-python.html) for using
>> Random Forest Classifer. The result looks working. From the sample code, it
>> looks like taking all attributes to train the model. But checking api (
>> http://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestClassifier.html
>> )
>> there is no description telling that what parameter can be used to ignore
>> some particular fileds/ attributes. So I would like to check if any
>> parameter can be used to ignore particular fields or attributes?
>> >
>> >Thanks
>> >
>>
>> >------------------------------------------------------------------------------
>> >Get 100% visibility into Java/.NET code with AppDynamics Lite!
>> >It's a free troubleshooting tool designed for production.
>> >Get down to code-level detail for bottlenecks, with <2% overhead.
>> >Download for free and get started troubleshooting in minutes.
>> >
>> http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
>> >_______________________________________________
>> >Scikit-learn-general mailing list
>> >[email protected]
>> >https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
>> >
>>
>>
>> --
>>
>> Roland
>> http://rolisz.ro/
>>
>>
>> ------------------------------------------------------------------------------
>> Get 100% visibility into Java/.NET code with AppDynamics Lite!
>> It's a free troubleshooting tool designed for production.
>> Get down to code-level detail for bottlenecks, with <2% overhead.
>> Download for free and get started troubleshooting in minutes.
>>
>> http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
>> _______________________________________________
>> Scikit-learn-general mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
>>
>
>
>
> ------------------------------------------------------------------------------
> Get 100% visibility into Java/.NET code with AppDynamics Lite!
> It's a free troubleshooting tool designed for production.
> Get down to code-level detail for bottlenecks, with <2% overhead.
> Download for free and get started troubleshooting in minutes.
> http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
> _______________________________________________
> Scikit-learn-general mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
>
>
------------------------------------------------------------------------------
Get 100% visibility into Java/.NET code with AppDynamics Lite!
It's a free troubleshooting tool designed for production.
Get down to code-level detail for bottlenecks, with <2% overhead.
Download for free and get started troubleshooting in minutes.
http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
