Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Plain View
Hadoop, mail # user - SVM implementaion


+
unmesha sreeveni 2013-11-08, 05:14
+
Mirko Kämpf 2013-11-08, 05:18
Copy link to this message
-
Re: SVM implementaion
unmesha sreeveni 2013-11-11, 06:22
can any one share a file with 1/-1 classification dataset..??
On Mon, Nov 11, 2013 at 11:26 AM, unmesha sreeveni <[EMAIL PROTECTED]>wrote:

> I am searching for numeric values loke Spam Classification..
>
>
> On Mon, Nov 11, 2013 at 11:21 AM, unmesha sreeveni <[EMAIL PROTECTED]>wrote:
>
>> it seems like all are image files...
>>
>>
>> On Sat, Nov 9, 2013 at 4:55 PM, unmesha sreeveni <[EMAIL PROTECTED]>wrote:
>>
>>> where to get the datasets?
>>> Whether in
>>> Active Competitions <http://www.kaggle.com/competitions>
>>>
>>>
>>> On Sat, Nov 9, 2013 at 4:51 PM, unmesha sreeveni <[EMAIL PROTECTED]>wrote:
>>>
>>>> normally in svm how are the training data classified in 1/0 class label?
>>>>
>>>> In 1/-1 classification we will be checking sign(decision function) ie ,
>>>> if the calculated decision function value is negative we will give -1
>>>> classlabel and if it is positive we will give 1 as class label.
>>>>
>>>> After finding that we will compare the class label with the class label
>>>> of input data...and find the accuracy/miss classification.
>>>>
>>>> What about in 1/0 classification???
>>>>
>>>>
>>>> On Sat, Nov 9, 2013 at 4:45 PM, unmesha sreeveni <[EMAIL PROTECTED]
>>>> > wrote:
>>>>
>>>>> 1/-1 is not available?
>>>>>
>>>>>
>>>>> On Fri, Nov 8, 2013 at 7:58 PM, Robin East <[EMAIL PROTECTED]>wrote:
>>>>>
>>>>>> You could try having a look at some of the competition data sets on
>>>>>> http://www.kaggle.com. Many of them are 1/0 classification problems
>>>>>>
>>>>>> Sent from my iPhone
>>>>>>
>>>>>> On 8 Nov 2013, at 10:10, unmesha sreeveni <[EMAIL PROTECTED]>
>>>>>> wrote:
>>>>>>
>>>>>> where can i get a large file for checking svm .with classes 1 and -1
>>>>>> around 1 GB file
>>>>>>
>>>>>>
>>>>>> On Fri, Nov 8, 2013 at 11:14 AM, unmesha sreeveni <
>>>>>> [EMAIL PROTECTED]> wrote:
>>>>>>
>>>>>>> thx Mirko Kampf
>>>>>>>
>>>>>>>
>>>>>>> On Fri, Nov 8, 2013 at 10:48 AM, Mirko Kämpf <[EMAIL PROTECTED]
>>>>>>> > wrote:
>>>>>>>
>>>>>>>> There are three Jira issues about it: Mahout-14<https://issues.apache.org/jira/browse/MAHOUT-14>
>>>>>>>>  and Mahout-334 <https://issues.apache.org/jira/browse/MAHOUT-334> have
>>>>>>>> been closed as won't fix. Mahout-232<https://issues.apache.org/jira/browse/MAHOUT-232> was
>>>>>>>> assigned later because some code was contributed early (2009) but it did
>>>>>>>> not work so it was not incorporated into Mahout. Since then Mahout has
>>>>>>>> changed, so porting the code would be difficult, and if you look at the
>>>>>>>> issue there is some disagreement about whether it approaches the problem in
>>>>>>>> the best way.
>>>>>>>>
>>>>>>>> There is some code for a cascadeSVM implementation<https://code.google.com/p/cascadesvm/> but
>>>>>>>> the training part - the hard part - was never published.
>>>>>>>>
>>>>>>>> There is a parallel SVM implementation<https://code.google.com/p/psvm/> that
>>>>>>>> runs on MPI rather than Hadoop.
>>>>>>>>
>>>>>>>> This San Francisco Meetup abstract<http://www.meetup.com/sfmachinelearning/events/116497192/> has
>>>>>>>> some discussion of current state of the art and issues for parallel SVM.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> 2013/11/8 unmesha sreeveni <[EMAIL PROTECTED]>
>>>>>>>>
>>>>>>>>> Do we have an SVM implementation?
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> *Thanks & Regards*
>>>>>>>>>
>>>>>>>>> Unmesha Sreeveni U.B
>>>>>>>>>
>>>>>>>>> *Junior Developer*
>>>>>>>>>
>>>>>>>>> *Amrita Center For Cyber Security *
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> * Amritapuri.www.amrita.edu/cyber/ <http://www.amrita.edu/cyber/>*
>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> *Thanks & Regards*
>>>>>>>
>>>>>>> Unmesha Sreeveni U.B
>>>>>>>
>>>>>>> *Junior Developer*
>>>>>>>
>>>>>>> *Amrita Center For Cyber Security *
>>>>>>>
>>>>>>>
>>>>>>> * Amritapuri.www.amrita.edu/cyber/ <http://www.amrita.edu/cyber/>*
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> *Thanks & Regards*
>>>>>>
>>>>>> Unmesha Sreeveni U.B
>>>>>>

*Thanks & Regards*

Unmesha Sreeveni U.B

*Junior Developer*

*Amrita Center For Cyber Security*
* Amritapuri.www.amrita.edu/cyber/ <http://www.amrita.edu/cyber/>*
+
unmesha sreeveni 2013-11-11, 05:51