You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@mahout.apache.org by Chameera Wijebandara <ch...@gmail.com> on 2014/01/24 05:00:19 UTC
Classify Handwritten Digits
Hi,
I am trying to classify handwritten digits using mahout classification. Any
suggestion to come up with good solution?
--
Thanks,
Chameera
Re: Classify Handwritten Digits
Posted by Chameera Wijebandara <ch...@gmail.com>.
Thanks Ted.
Thanks,
Chameera
On Fri, Jan 24, 2014 at 11:12 PM, Ted Dunning <te...@gmail.com> wrote:
> You can also put out lots of clusters and use cluster membership as the
> features for a classifier.
>
> There was a discussion here (or possibly on the dev@mahout list) on this
> topic several weeks ago. Search the archives for "iris" and my name.
>
>
>
> On Fri, Jan 24, 2014 at 8:46 AM, Angus Macnab <angus.macnab@gmail.com
> >wrote:
>
> > You can do supervised learning by outputing the clusters and labeling
> them
> > 0-9.
> >
> > > On Jan 23, 2014, at 10:34 PM, Tharindu Rusira <
> tharindurusira@gmail.com>
> > wrote:
> > >
> > > On Fri, Jan 24, 2014 at 9:50 AM, Angus Macnab <angus.macnab@gmail.com
> > >wrote:
> > >
> > >> This is a pretty classic machine learning problem and can be handled
> > with
> > >> several different algorithms. Logistic regression is the obvious
> > choice,
> > >> but clustering algorithms will work fine also. Just decompose the
> > pixels
> > >> into a really long vector and train your algorithm with the
> input-output
> > >> pairs. You can get 100% accuracy on this pretty easily if you are
> > careful
> > >> with your bias-variance decomposition. This is a fun one for neural
> > >> networks too!
> > >>
> > >> Essentially any machine learning book will delve into greater detail
> on
> > >> this as the US postal digit data has been around for a long time. I
> > think
> > >> Kaggle even had this as a training exercise for a while, so there's
> > >> probably a ton of discussion of various methods and algorithms on
> their
> > >> message boards.
> > >>
> > >> For kicks why don't you compare k-means clustering to logistic
> > regression
> > >> using Mahout?
> > > Hi Angus, Chameera's requirement is to classify handwritten digits, so
> > > could you please explain how could K-means clustering be helpful in
> this
> > > scenario? Of course it would find different clusters but this is still
> a
> > > classification problem. Please correct me if I'm wrong.
> > >
> > > Thanks,
> > >
> > >
> > >>
> > >> -Angus
> > >>
> > >>
> > >>
> > >>
> > >> On Thu, Jan 23, 2014 at 8:00 PM, Chameera Wijebandara <
> > >> chameerawijebandara@gmail.com> wrote:
> > >>
> > >>> Hi,
> > >>>
> > >>> I am trying to classify handwritten digits using mahout
> classification.
> > >> Any
> > >>> suggestion to come up with good solution?
> > >>>
> > >>> --
> > >>> Thanks,
> > >>> Chameera
> > >
> > >
> > >
> > > --
> > > M.P. Tharindu Rusira Kumara
> > >
> > > Department of Computer Science and Engineering,
> > > University of Moratuwa,
> > > Sri Lanka.
> > > +94757033733
> > > www.tharindu-rusira.blogspot.com
> >
>
Re: Classify Handwritten Digits
Posted by Ted Dunning <te...@gmail.com>.
You can also put out lots of clusters and use cluster membership as the
features for a classifier.
There was a discussion here (or possibly on the dev@mahout list) on this
topic several weeks ago. Search the archives for "iris" and my name.
On Fri, Jan 24, 2014 at 8:46 AM, Angus Macnab <an...@gmail.com>wrote:
> You can do supervised learning by outputing the clusters and labeling them
> 0-9.
>
> > On Jan 23, 2014, at 10:34 PM, Tharindu Rusira <th...@gmail.com>
> wrote:
> >
> > On Fri, Jan 24, 2014 at 9:50 AM, Angus Macnab <angus.macnab@gmail.com
> >wrote:
> >
> >> This is a pretty classic machine learning problem and can be handled
> with
> >> several different algorithms. Logistic regression is the obvious
> choice,
> >> but clustering algorithms will work fine also. Just decompose the
> pixels
> >> into a really long vector and train your algorithm with the input-output
> >> pairs. You can get 100% accuracy on this pretty easily if you are
> careful
> >> with your bias-variance decomposition. This is a fun one for neural
> >> networks too!
> >>
> >> Essentially any machine learning book will delve into greater detail on
> >> this as the US postal digit data has been around for a long time. I
> think
> >> Kaggle even had this as a training exercise for a while, so there's
> >> probably a ton of discussion of various methods and algorithms on their
> >> message boards.
> >>
> >> For kicks why don't you compare k-means clustering to logistic
> regression
> >> using Mahout?
> > Hi Angus, Chameera's requirement is to classify handwritten digits, so
> > could you please explain how could K-means clustering be helpful in this
> > scenario? Of course it would find different clusters but this is still a
> > classification problem. Please correct me if I'm wrong.
> >
> > Thanks,
> >
> >
> >>
> >> -Angus
> >>
> >>
> >>
> >>
> >> On Thu, Jan 23, 2014 at 8:00 PM, Chameera Wijebandara <
> >> chameerawijebandara@gmail.com> wrote:
> >>
> >>> Hi,
> >>>
> >>> I am trying to classify handwritten digits using mahout classification.
> >> Any
> >>> suggestion to come up with good solution?
> >>>
> >>> --
> >>> Thanks,
> >>> Chameera
> >
> >
> >
> > --
> > M.P. Tharindu Rusira Kumara
> >
> > Department of Computer Science and Engineering,
> > University of Moratuwa,
> > Sri Lanka.
> > +94757033733
> > www.tharindu-rusira.blogspot.com
>
Re: Classify Handwritten Digits
Posted by Angus Macnab <an...@gmail.com>.
You can do supervised learning by outputing the clusters and labeling them 0-9.
> On Jan 23, 2014, at 10:34 PM, Tharindu Rusira <th...@gmail.com> wrote:
>
> On Fri, Jan 24, 2014 at 9:50 AM, Angus Macnab <an...@gmail.com>wrote:
>
>> This is a pretty classic machine learning problem and can be handled with
>> several different algorithms. Logistic regression is the obvious choice,
>> but clustering algorithms will work fine also. Just decompose the pixels
>> into a really long vector and train your algorithm with the input-output
>> pairs. You can get 100% accuracy on this pretty easily if you are careful
>> with your bias-variance decomposition. This is a fun one for neural
>> networks too!
>>
>> Essentially any machine learning book will delve into greater detail on
>> this as the US postal digit data has been around for a long time. I think
>> Kaggle even had this as a training exercise for a while, so there's
>> probably a ton of discussion of various methods and algorithms on their
>> message boards.
>>
>> For kicks why don't you compare k-means clustering to logistic regression
>> using Mahout?
> Hi Angus, Chameera's requirement is to classify handwritten digits, so
> could you please explain how could K-means clustering be helpful in this
> scenario? Of course it would find different clusters but this is still a
> classification problem. Please correct me if I'm wrong.
>
> Thanks,
>
>
>>
>> -Angus
>>
>>
>>
>>
>> On Thu, Jan 23, 2014 at 8:00 PM, Chameera Wijebandara <
>> chameerawijebandara@gmail.com> wrote:
>>
>>> Hi,
>>>
>>> I am trying to classify handwritten digits using mahout classification.
>> Any
>>> suggestion to come up with good solution?
>>>
>>> --
>>> Thanks,
>>> Chameera
>
>
>
> --
> M.P. Tharindu Rusira Kumara
>
> Department of Computer Science and Engineering,
> University of Moratuwa,
> Sri Lanka.
> +94757033733
> www.tharindu-rusira.blogspot.com
Re: Classify Handwritten Digits
Posted by Tharindu Rusira <th...@gmail.com>.
On Fri, Jan 24, 2014 at 9:50 AM, Angus Macnab <an...@gmail.com>wrote:
> This is a pretty classic machine learning problem and can be handled with
> several different algorithms. Logistic regression is the obvious choice,
> but clustering algorithms will work fine also. Just decompose the pixels
> into a really long vector and train your algorithm with the input-output
> pairs. You can get 100% accuracy on this pretty easily if you are careful
> with your bias-variance decomposition. This is a fun one for neural
> networks too!
>
> Essentially any machine learning book will delve into greater detail on
> this as the US postal digit data has been around for a long time. I think
> Kaggle even had this as a training exercise for a while, so there's
> probably a ton of discussion of various methods and algorithms on their
> message boards.
>
> For kicks why don't you compare k-means clustering to logistic regression
> using Mahout?
>
Hi Angus, Chameera's requirement is to classify handwritten digits, so
could you please explain how could K-means clustering be helpful in this
scenario? Of course it would find different clusters but this is still a
classification problem. Please correct me if I'm wrong.
Thanks,
>
> -Angus
>
>
>
>
> On Thu, Jan 23, 2014 at 8:00 PM, Chameera Wijebandara <
> chameerawijebandara@gmail.com> wrote:
>
> > Hi,
> >
> > I am trying to classify handwritten digits using mahout classification.
> Any
> > suggestion to come up with good solution?
> >
> > --
> > Thanks,
> > Chameera
> >
>
--
M.P. Tharindu Rusira Kumara
Department of Computer Science and Engineering,
University of Moratuwa,
Sri Lanka.
+94757033733
www.tharindu-rusira.blogspot.com
Re: Classify Handwritten Digits
Posted by Chameera Wijebandara <ch...@gmail.com>.
Hi Angus,
Thank you for your detailed explanation.
I have already implemented neural network approach and had success rate.
And at the moment the requirement of doing this is getting familiar with
mahout. Are there any tutorial about K -means or logistic regression with
mahout?
Thank,
Chameera.
On Fri, Jan 24, 2014 at 9:50 AM, Angus Macnab <an...@gmail.com>wrote:
> This is a pretty classic machine learning problem and can be handled with
> several different algorithms. Logistic regression is the obvious choice,
> but clustering algorithms will work fine also. Just decompose the pixels
> into a really long vector and train your algorithm with the input-output
> pairs. You can get 100% accuracy on this pretty easily if you are careful
> with your bias-variance decomposition. This is a fun one for neural
> networks too!
>
> Essentially any machine learning book will delve into greater detail on
> this as the US postal digit data has been around for a long time. I think
> Kaggle even had this as a training exercise for a while, so there's
> probably a ton of discussion of various methods and algorithms on their
> message boards.
>
> For kicks why don't you compare k-means clustering to logistic regression
> using Mahout?
>
> -Angus
>
>
>
>
> On Thu, Jan 23, 2014 at 8:00 PM, Chameera Wijebandara <
> chameerawijebandara@gmail.com> wrote:
>
> > Hi,
> >
> > I am trying to classify handwritten digits using mahout classification.
> Any
> > suggestion to come up with good solution?
> >
> > --
> > Thanks,
> > Chameera
> >
>
--
Thanks,
Chameera
Re: Classify Handwritten Digits
Posted by Angus Macnab <an...@gmail.com>.
This is a pretty classic machine learning problem and can be handled with
several different algorithms. Logistic regression is the obvious choice,
but clustering algorithms will work fine also. Just decompose the pixels
into a really long vector and train your algorithm with the input-output
pairs. You can get 100% accuracy on this pretty easily if you are careful
with your bias-variance decomposition. This is a fun one for neural
networks too!
Essentially any machine learning book will delve into greater detail on
this as the US postal digit data has been around for a long time. I think
Kaggle even had this as a training exercise for a while, so there's
probably a ton of discussion of various methods and algorithms on their
message boards.
For kicks why don't you compare k-means clustering to logistic regression
using Mahout?
-Angus
On Thu, Jan 23, 2014 at 8:00 PM, Chameera Wijebandara <
chameerawijebandara@gmail.com> wrote:
> Hi,
>
> I am trying to classify handwritten digits using mahout classification. Any
> suggestion to come up with good solution?
>
> --
> Thanks,
> Chameera
>
Re: Classify Handwritten Digits
Posted by Chameera Wijebandara <ch...@gmail.com>.
Hi Koji
Thanks.
Ill try it
Thanks,
Chameera
On Fri, Jan 24, 2014 at 3:11 PM, Koji Sekiguchi <ko...@r.email.ne.jp> wrote:
> (14/01/24 13:00), Chameera Wijebandara wrote:
>
>> Hi,
>>
>> I am trying to classify handwritten digits using mahout classification.
>> Any
>> suggestion to come up with good solution?
>>
>
> In the Mahout training course, we have an exercise of handwritten character
> recognition with Mahout:
>
> http://soleami.com/blog/mahout-and-machine-learning-
> training-course-is-here.html
>
> In the exercise, we use Mahout's random forest.
>
> For those who are interested, there is a demo:
>
> http://rondhuit-demo.com/p/hwcr/recognition
>
> How to use demo:
>
> 1. Using a mouse, draw a digit ('0', '1', ..., '9') in the smaller box
> (HTML5 Canvas)
> 2. Click [Classify] button (Do not click more than once, please!)
> 3. Wait 15 seconds, you'll get the result in the larger box
>
> Note that we use the nearest neighbor method (not random forest) in this
> demo.
>
> Hope this helps.
>
> koji
> --
> http://soleami.com/blog/mahout-and-machine-learning-
> training-course-is-here.html
>
Re: Classify Handwritten Digits
Posted by Koji Sekiguchi <ko...@r.email.ne.jp>.
(14/01/24 13:00), Chameera Wijebandara wrote:
> Hi,
>
> I am trying to classify handwritten digits using mahout classification. Any
> suggestion to come up with good solution?
In the Mahout training course, we have an exercise of handwritten character
recognition with Mahout:
http://soleami.com/blog/mahout-and-machine-learning-training-course-is-here.html
In the exercise, we use Mahout's random forest.
For those who are interested, there is a demo:
http://rondhuit-demo.com/p/hwcr/recognition
How to use demo:
1. Using a mouse, draw a digit ('0', '1', ..., '9') in the smaller box (HTML5 Canvas)
2. Click [Classify] button (Do not click more than once, please!)
3. Wait 15 seconds, you'll get the result in the larger box
Note that we use the nearest neighbor method (not random forest) in this demo.
Hope this helps.
koji
--
http://soleami.com/blog/mahout-and-machine-learning-training-course-is-here.html