Re: [Dev] GSoC project for building an ensemble of models

[Supun Sethunga]

Hi Anurag,

Please submit your final proposal to GSoC, before the deadline.

Regards,
Supun

On Tue, Mar 22, 2016 at 11:12 PM, Supun Sethunga  wrote:

> Hi Anurag,
>
> Looks good! May be you can also improve the diagram to reflect the
> description of the three methods (which is right above the diagram).
>
> Anyway, please create the final proposal and submit before Friday, 25th.
>
> Regards,
> Supun
>
> On Tue, Mar 22, 2016 at 1:52 PM, Anurag Kashyap 
> wrote:
>
>> Hi,
>>  I have made a few changes to the proposal and elaborated a few points.
>> Kindly give your valuable feedback.
>>
>> https://docs.google.com/document/d/1L132ApiOm4m0DbyVZ-TYCedsRrt7oE_cgOzAuScvFiQ/edit?usp=sharing
>>
>> Regards,
>> Anurag Kashyap
>>
>> On Mon, Mar 21, 2016 at 4:43 PM, Supun Sethunga  wrote:
>>
>>> Hi Anurag,
>>>
>>> We have added a few comments. Can you please fix those and share the
>>> updated one?
>>>
>>> Thanks,
>>> Supun
>>>
>>> On Mon, Mar 21, 2016 at 1:29 PM, Nirmal Fernando 
>>> wrote:
>>>
 Thanks.

 On Mon, Mar 21, 2016 at 1:27 PM, Anurag Kashyap 
 wrote:

> Hi,
>  I've changed the settings. You should be able to comment or edit the
> proposa
> l
> now.
> Thanks
>
> On Mon, Mar 21, 2016 at 3:52 PM, Nirmal Fernando 
> wrote:
>
>> Hi Anurag,
>>
>> Thanks for your proposal. Can you please give us 'commenting' right?
>>
>> On Mon, Mar 21, 2016 at 1:03 PM, Anurag Kashyap > > wrote:
>>
>>> Hi,
>>> Thank you so much for the valuable input. This
>>> 
>>> is a tentative draft of my proposal.
>>>
>>> Regards,
>>> Anurag
>>>
>>> On Mon, Mar 21, 2016 at 1:03 PM, Supun Sethunga 
>>> wrote:
>>>
 Hi Anurag,

 Have you started on the project proposal? If so, could you please
 share it (the draft) with us?

 Thanks,

 On Mon, Mar 21, 2016 at 10:03 AM, Nirmal Fernando 
 wrote:

>
>
> On Mon, Mar 21, 2016 at 10:01 AM, Supun Sethunga 
> wrote:
>
>> Hi Anurag,
>>
>> Can you help me figure out the significance of the
>>> serialVersionUID field?
>>
>>
>> All the algorithm classes including the one you mentioned
>> implement the Serializable interface (Because these generated models 
>> can be
>> exported and imported as java objects). "serialVersionUID" is
>> used when serializing and deserializing the objects. If we do not 
>> define a "serialVersionUID",
>> then java runtime will generate them during serialization and
>> deserialization. And if these two serialVersionUID's does not
>> matches, then the deserialization fails. As a result, the java 
>> object will
>> not be readable. You can check [1] for further details.
>>
>> Also, I wanted to know if I should include personal details too
>>> in the proposal- such as my university and where I'm currently 
>>> working.
>>
>> You may include. You can check out a sample gsoc proposal by
>> Nirmal (nir...@wso2.com) at [2].
>>
>> BTW, please CC to WSO2 Developers mailing list (dev@wso2.org)
>> for all project-related discussions.
>>
>
> You may have to subscribe to dev@, please refer to
> http://wso2.com/mail/
>
>>
>> [1]
>> http://javarevisited.blogspot.com/2014/05/why-use-serialversionuid-inside-serializable-class-in-java.html
>> [2] http://nirmalfdo.blogspot.com/search/label/GSoC-2011
>>
>> Regards
>> Supun
>>
>>
>> On Thu, Mar 17, 2016 at 6:26 PM, Anurag Kashyap <
>> anurag1...@gmail.com> wrote:
>>
>>> Hi Supun,
>>>  I was going through this
>>> 
>>> link. Can you help me figure out the significance of the 
>>> serialVersionUID
>>> field? If that field is necessary then anonymous classes will have 
>>> to be
>>> implemented instead of lambda expressions in several places. Also, 
>>> I wanted
>>> to know if I should include personal details too in the proposal- 
>>> such as
>>> my university and where I'm currently working.
>>>
>>>
>>>

Re: [Dev] GSoC project for building an ensemble of models

[Supun Sethunga]

Hi Anurag,

Looks good! May be you can also improve the diagram to reflect the
description of the three methods (which is right above the diagram).

Anyway, please create the final proposal and submit before Friday, 25th.

Regards,
Supun

On Tue, Mar 22, 2016 at 1:52 PM, Anurag Kashyap 
wrote:

> Hi,
>  I have made a few changes to the proposal and elaborated a few points.
> Kindly give your valuable feedback.
>
> https://docs.google.com/document/d/1L132ApiOm4m0DbyVZ-TYCedsRrt7oE_cgOzAuScvFiQ/edit?usp=sharing
>
> Regards,
> Anurag Kashyap
>
> On Mon, Mar 21, 2016 at 4:43 PM, Supun Sethunga  wrote:
>
>> Hi Anurag,
>>
>> We have added a few comments. Can you please fix those and share the
>> updated one?
>>
>> Thanks,
>> Supun
>>
>> On Mon, Mar 21, 2016 at 1:29 PM, Nirmal Fernando  wrote:
>>
>>> Thanks.
>>>
>>> On Mon, Mar 21, 2016 at 1:27 PM, Anurag Kashyap 
>>> wrote:
>>>
 Hi,
  I've changed the settings. You should be able to comment or edit the
 proposa
 l
 now.
 Thanks

 On Mon, Mar 21, 2016 at 3:52 PM, Nirmal Fernando 
 wrote:

> Hi Anurag,
>
> Thanks for your proposal. Can you please give us 'commenting' right?
>
> On Mon, Mar 21, 2016 at 1:03 PM, Anurag Kashyap 
> wrote:
>
>> Hi,
>> Thank you so much for the valuable input. This
>> 
>> is a tentative draft of my proposal.
>>
>> Regards,
>> Anurag
>>
>> On Mon, Mar 21, 2016 at 1:03 PM, Supun Sethunga 
>> wrote:
>>
>>> Hi Anurag,
>>>
>>> Have you started on the project proposal? If so, could you please
>>> share it (the draft) with us?
>>>
>>> Thanks,
>>>
>>> On Mon, Mar 21, 2016 at 10:03 AM, Nirmal Fernando 
>>> wrote:
>>>


 On Mon, Mar 21, 2016 at 10:01 AM, Supun Sethunga 
 wrote:

> Hi Anurag,
>
> Can you help me figure out the significance of the
>> serialVersionUID field?
>
>
> All the algorithm classes including the one you mentioned
> implement the Serializable interface (Because these generated models 
> can be
> exported and imported as java objects). "serialVersionUID" is
> used when serializing and deserializing the objects. If we do not 
> define a "serialVersionUID",
> then java runtime will generate them during serialization and
> deserialization. And if these two serialVersionUID's does not
> matches, then the deserialization fails. As a result, the java object 
> will
> not be readable. You can check [1] for further details.
>
> Also, I wanted to know if I should include personal details too in
>> the proposal- such as my university and where I'm currently working.
>
> You may include. You can check out a sample gsoc proposal by
> Nirmal (nir...@wso2.com) at [2].
>
> BTW, please CC to WSO2 Developers mailing list (dev@wso2.org) for
> all project-related discussions.
>

 You may have to subscribe to dev@, please refer to
 http://wso2.com/mail/

>
> [1]
> http://javarevisited.blogspot.com/2014/05/why-use-serialversionuid-inside-serializable-class-in-java.html
> [2] http://nirmalfdo.blogspot.com/search/label/GSoC-2011
>
> Regards
> Supun
>
>
> On Thu, Mar 17, 2016 at 6:26 PM, Anurag Kashyap <
> anurag1...@gmail.com> wrote:
>
>> Hi Supun,
>>  I was going through this
>> 
>> link. Can you help me figure out the significance of the 
>> serialVersionUID
>> field? If that field is necessary then anonymous classes will have 
>> to be
>> implemented instead of lambda expressions in several places. Also, I 
>> wanted
>> to know if I should include personal details too in the proposal- 
>> such as
>> my university and where I'm currently working.
>>
>>
>>
>>
>>
>>
>>
>> On Mon, Mar 14, 2016 at 5:57 PM, Supun Sethunga 
>> wrote:
>>
>>> Hi Anurag,
>>>
>>> Yes, the master branch is using java 8, hence you can use lambda
>>> expressions.
>>>
>>> Regards,
>>> Supun

Re: [Dev] GSoC project for building an ensemble of models

[Supun Sethunga]

Hi Anurag,

We have added a few comments. Can you please fix those and share the
updated one?

Thanks,
Supun

On Mon, Mar 21, 2016 at 1:29 PM, Nirmal Fernando  wrote:

> Thanks.
>
> On Mon, Mar 21, 2016 at 1:27 PM, Anurag Kashyap 
> wrote:
>
>> Hi,
>>  I've changed the settings. You should be able to comment or edit the
>> proposa
>> l
>> now.
>> Thanks
>>
>> On Mon, Mar 21, 2016 at 3:52 PM, Nirmal Fernando  wrote:
>>
>>> Hi Anurag,
>>>
>>> Thanks for your proposal. Can you please give us 'commenting' right?
>>>
>>> On Mon, Mar 21, 2016 at 1:03 PM, Anurag Kashyap 
>>> wrote:
>>>
 Hi,
 Thank you so much for the valuable input. This
 
 is a tentative draft of my proposal.

 Regards,
 Anurag

 On Mon, Mar 21, 2016 at 1:03 PM, Supun Sethunga 
 wrote:

> Hi Anurag,
>
> Have you started on the project proposal? If so, could you please
> share it (the draft) with us?
>
> Thanks,
>
> On Mon, Mar 21, 2016 at 10:03 AM, Nirmal Fernando 
> wrote:
>
>>
>>
>> On Mon, Mar 21, 2016 at 10:01 AM, Supun Sethunga 
>> wrote:
>>
>>> Hi Anurag,
>>>
>>> Can you help me figure out the significance of the serialVersionUID
 field?
>>>
>>>
>>> All the algorithm classes including the one you mentioned
>>> implement the Serializable interface (Because these generated models 
>>> can be
>>> exported and imported as java objects). "serialVersionUID" is used
>>> when serializing and deserializing the objects. If we do not define a 
>>> "serialVersionUID",
>>> then java runtime will generate them during serialization and
>>> deserialization. And if these two serialVersionUID's does not
>>> matches, then the deserialization fails. As a result, the java object 
>>> will
>>> not be readable. You can check [1] for further details.
>>>
>>> Also, I wanted to know if I should include personal details too in
 the proposal- such as my university and where I'm currently working.
>>>
>>> You may include. You can check out a sample gsoc proposal by Nirmal (
>>> nir...@wso2.com) at [2].
>>>
>>> BTW, please CC to WSO2 Developers mailing list (dev@wso2.org) for
>>> all project-related discussions.
>>>
>>
>> You may have to subscribe to dev@, please refer to
>> http://wso2.com/mail/
>>
>>>
>>> [1]
>>> http://javarevisited.blogspot.com/2014/05/why-use-serialversionuid-inside-serializable-class-in-java.html
>>> [2] http://nirmalfdo.blogspot.com/search/label/GSoC-2011
>>>
>>> Regards
>>> Supun
>>>
>>>
>>> On Thu, Mar 17, 2016 at 6:26 PM, Anurag Kashyap <
>>> anurag1...@gmail.com> wrote:
>>>
 Hi Supun,
  I was going through this
 
 link. Can you help me figure out the significance of the 
 serialVersionUID
 field? If that field is necessary then anonymous classes will have to 
 be
 implemented instead of lambda expressions in several places. Also, I 
 wanted
 to know if I should include personal details too in the proposal- such 
 as
 my university and where I'm currently working.







 On Mon, Mar 14, 2016 at 5:57 PM, Supun Sethunga 
 wrote:

> Hi Anurag,
>
> Yes, the master branch is using java 8, hence you can use lambda
> expressions.
>
> Regards,
> Supun
>
> On Mon, Mar 14, 2016 at 3:24 PM, Anurag Kashyap <
> anurag1...@gmail.com> wrote:
>
>> Hi Supun,
>> One more question. I couldn't find if the version of Java being
>> used is Java 8 or a previous version. If it is Java 8, Lambda 
>> calculus can
>> be used to make some things easier, especially while working with 
>> spark.
>> Thanks and Regards,
>> Anurag Kashyap
>>
>>
>> On Mon, Mar 7, 2016 at 8:05 PM, Supun Sethunga 
>> wrote:
>>
>>> [Adding WSO2 Developers mailing list]
>>>
>>> Hi Anurag,
>>>
>>> Please find my comments inline.
>>>
>>> First, which boosting model are we going to target-Ada boosting
 or some other model, or various models?
>>>
>>> You can pick any 

Re: [Dev] GSoC project for building an ensemble of models

[Supun Sethunga]

Hi Anurag,

Have you started on the project proposal? If so, could you please share it
(the draft) with us?

Thanks,

On Mon, Mar 21, 2016 at 10:03 AM, Nirmal Fernando  wrote:

>
>
> On Mon, Mar 21, 2016 at 10:01 AM, Supun Sethunga  wrote:
>
>> Hi Anurag,
>>
>> Can you help me figure out the significance of the serialVersionUID field?
>>
>>
>> All the algorithm classes including the one you mentioned implement the
>> Serializable interface (Because these generated models can be exported and
>> imported as java objects). "serialVersionUID" is used when serializing
>> and deserializing the objects. If we do not define a "serialVersionUID",
>> then java runtime will generate them during serialization and
>> deserialization. And if these two serialVersionUID's does not matches,
>> then the deserialization fails. As a result, the java object will not be
>> readable. You can check [1] for further details.
>>
>> Also, I wanted to know if I should include personal details too in the
>>> proposal- such as my university and where I'm currently working.
>>
>> You may include. You can check out a sample gsoc proposal by Nirmal (
>> nir...@wso2.com) at [2].
>>
>> BTW, please CC to WSO2 Developers mailing list (dev@wso2.org) for all
>> project-related discussions.
>>
>
> You may have to subscribe to dev@, please refer to http://wso2.com/mail/
>
>>
>> [1]
>> http://javarevisited.blogspot.com/2014/05/why-use-serialversionuid-inside-serializable-class-in-java.html
>> [2] http://nirmalfdo.blogspot.com/search/label/GSoC-2011
>>
>> Regards
>> Supun
>>
>>
>> On Thu, Mar 17, 2016 at 6:26 PM, Anurag Kashyap 
>> wrote:
>>
>>> Hi Supun,
>>>  I was going through this
>>> 
>>> link. Can you help me figure out the significance of the serialVersionUID
>>> field? If that field is necessary then anonymous classes will have to be
>>> implemented instead of lambda expressions in several places. Also, I wanted
>>> to know if I should include personal details too in the proposal- such as
>>> my university and where I'm currently working.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> On Mon, Mar 14, 2016 at 5:57 PM, Supun Sethunga  wrote:
>>>
 Hi Anurag,

 Yes, the master branch is using java 8, hence you can use lambda
 expressions.

 Regards,
 Supun

 On Mon, Mar 14, 2016 at 3:24 PM, Anurag Kashyap 
 wrote:

> Hi Supun,
> One more question. I couldn't find if the version of Java being used
> is Java 8 or a previous version. If it is Java 8, Lambda calculus can be
> used to make some things easier, especially while working with spark.
> Thanks and Regards,
> Anurag Kashyap
>
>
> On Mon, Mar 7, 2016 at 8:05 PM, Supun Sethunga 
> wrote:
>
>> [Adding WSO2 Developers mailing list]
>>
>> Hi Anurag,
>>
>> Please find my comments inline.
>>
>> First, which boosting model are we going to target-Ada boosting or
>>> some other model, or various models?
>>
>> You can pick any method as you wish.
>>
>>
>> Second, what method would you suggest I use to combine the results of
>>> the models in stacking- is it going to be a single layer LR model?
>>
>> This has to be a user input (i.e. user should be able to pick any
>> algorithm for the base-learners as well as for the meta-learner).
>>
>> Regards,
>> Supun
>>
>> On Mon, Mar 7, 2016 at 5:15 PM, Anurag Kashyap 
>> wrote:
>>
>>> Hi Supun,
>>>  I've begun reading the source rom link 3. I had a couple of
>>> questions. First, which boosting model are we going to target-Ada 
>>> boosting
>>> or some other model, or various models? Second, what method would you
>>> suggest I use to combine the results of the models in stacking- is it 
>>> going
>>> to be a single layer LR model?
>>>
>>>
>>>
>>> On Sun, Mar 6, 2016 at 6:25 PM, Supun Sethunga 
>>> wrote:
>>>
 Hi Anurag,

 Thank you for your interest in the above project. You can start by
 getting to know the Machine Learner product by downloading it and 
 running
 it (Please use link [1] to download). Official documentation [2] will 
 help
 you on using the product. You can find the source code base at [3] and 
 [4].
 As we have mentioned in the project proposal as well, the main
 objective is to integrate ensemble support for the existing flow of the
 WSO2 Machine Learner. We are focusing on the three methods: Bagging,
 Boosting and Stacking.

 Please feel free to raise, if you have any questions or any unclear

Re: [Dev] GSoC project for building an ensemble of models

[Nirmal Fernando]

On Mon, Mar 21, 2016 at 10:01 AM, Supun Sethunga  wrote:

> Hi Anurag,
>
> Can you help me figure out the significance of the serialVersionUID field?
>
>
> All the algorithm classes including the one you mentioned implement the
> Serializable interface (Because these generated models can be exported and
> imported as java objects). "serialVersionUID" is used when serializing
> and deserializing the objects. If we do not define a "serialVersionUID",
> then java runtime will generate them during serialization and
> deserialization. And if these two serialVersionUID's does not matches,
> then the deserialization fails. As a result, the java object will not be
> readable. You can check [1] for further details.
>
> Also, I wanted to know if I should include personal details too in the
>> proposal- such as my university and where I'm currently working.
>
> You may include. You can check out a sample gsoc proposal by Nirmal (
> nir...@wso2.com) at [2].
>
> BTW, please CC to WSO2 Developers mailing list (dev@wso2.org) for all
> project-related discussions.
>

You may have to subscribe to dev@, please refer to http://wso2.com/mail/

>
> [1]
> http://javarevisited.blogspot.com/2014/05/why-use-serialversionuid-inside-serializable-class-in-java.html
> [2] http://nirmalfdo.blogspot.com/search/label/GSoC-2011
>
> Regards
> Supun
>
>
> On Thu, Mar 17, 2016 at 6:26 PM, Anurag Kashyap 
> wrote:
>
>> Hi Supun,
>>  I was going through this
>> 
>> link. Can you help me figure out the significance of the serialVersionUID
>> field? If that field is necessary then anonymous classes will have to be
>> implemented instead of lambda expressions in several places. Also, I wanted
>> to know if I should include personal details too in the proposal- such as
>> my university and where I'm currently working.
>>
>>
>>
>>
>>
>>
>>
>> On Mon, Mar 14, 2016 at 5:57 PM, Supun Sethunga  wrote:
>>
>>> Hi Anurag,
>>>
>>> Yes, the master branch is using java 8, hence you can use lambda
>>> expressions.
>>>
>>> Regards,
>>> Supun
>>>
>>> On Mon, Mar 14, 2016 at 3:24 PM, Anurag Kashyap 
>>> wrote:
>>>
 Hi Supun,
 One more question. I couldn't find if the version of Java being used is
 Java 8 or a previous version. If it is Java 8, Lambda calculus can be used
 to make some things easier, especially while working with spark.
 Thanks and Regards,
 Anurag Kashyap


 On Mon, Mar 7, 2016 at 8:05 PM, Supun Sethunga  wrote:

> [Adding WSO2 Developers mailing list]
>
> Hi Anurag,
>
> Please find my comments inline.
>
> First, which boosting model are we going to target-Ada boosting or
>> some other model, or various models?
>
> You can pick any method as you wish.
>
>
> Second, what method would you suggest I use to combine the results of
>> the models in stacking- is it going to be a single layer LR model?
>
> This has to be a user input (i.e. user should be able to pick any
> algorithm for the base-learners as well as for the meta-learner).
>
> Regards,
> Supun
>
> On Mon, Mar 7, 2016 at 5:15 PM, Anurag Kashyap 
> wrote:
>
>> Hi Supun,
>>  I've begun reading the source rom link 3. I had a couple of
>> questions. First, which boosting model are we going to target-Ada 
>> boosting
>> or some other model, or various models? Second, what method would you
>> suggest I use to combine the results of the models in stacking- is it 
>> going
>> to be a single layer LR model?
>>
>>
>>
>> On Sun, Mar 6, 2016 at 6:25 PM, Supun Sethunga 
>> wrote:
>>
>>> Hi Anurag,
>>>
>>> Thank you for your interest in the above project. You can start by
>>> getting to know the Machine Learner product by downloading it and 
>>> running
>>> it (Please use link [1] to download). Official documentation [2] will 
>>> help
>>> you on using the product. You can find the source code base at [3] and 
>>> [4].
>>> As we have mentioned in the project proposal as well, the main
>>> objective is to integrate ensemble support for the existing flow of the
>>> WSO2 Machine Learner. We are focusing on the three methods: Bagging,
>>> Boosting and Stacking.
>>>
>>> Please feel free to raise, if you have any questions or any unclear
>>> points.
>>>
>>> [1] http://wso2.com/products/machine-learner/
>>> [2] https://docs.wso2.com/display/ML100/Introducing+Machine+Learner
>>> [3] https://github.com/wso2/carbon-ml
>>> [4] https://github.com/wso2/product-ml
>>>
>>> Regards,
>>> Supun
>>>
>>> On Sun, Mar 6, 2016 at 12:08 PM, 

Re: [Dev] GSoC project for building an ensemble of models

[Supun Sethunga]

Hi Anurag,

Can you help me figure out the significance of the serialVersionUID field?


All the algorithm classes including the one you mentioned implement the
Serializable interface (Because these generated models can be exported and
imported as java objects). "serialVersionUID" is used when serializing and
deserializing the objects. If we do not define a "serialVersionUID", then
java runtime will generate them during serialization and deserialization.
And if these two serialVersionUID's does not matches, then the
deserialization fails. As a result, the java object will not be readable.
You can check [1] for further details.

Also, I wanted to know if I should include personal details too in the
> proposal- such as my university and where I'm currently working.

You may include. You can check out a sample gsoc proposal by Nirmal (
nir...@wso2.com) at [2].

BTW, please CC to WSO2 Developers mailing list (dev@wso2.org) for all
project-related discussions.

[1]
http://javarevisited.blogspot.com/2014/05/why-use-serialversionuid-inside-serializable-class-in-java.html
[2] http://nirmalfdo.blogspot.com/search/label/GSoC-2011

Regards
Supun


On Thu, Mar 17, 2016 at 6:26 PM, Anurag Kashyap 
wrote:

> Hi Supun,
>  I was going through this
> 
> link. Can you help me figure out the significance of the serialVersionUID
> field? If that field is necessary then anonymous classes will have to be
> implemented instead of lambda expressions in several places. Also, I wanted
> to know if I should include personal details too in the proposal- such as
> my university and where I'm currently working.
>
>
>
>
>
>
>
> On Mon, Mar 14, 2016 at 5:57 PM, Supun Sethunga  wrote:
>
>> Hi Anurag,
>>
>> Yes, the master branch is using java 8, hence you can use lambda
>> expressions.
>>
>> Regards,
>> Supun
>>
>> On Mon, Mar 14, 2016 at 3:24 PM, Anurag Kashyap 
>> wrote:
>>
>>> Hi Supun,
>>> One more question. I couldn't find if the version of Java being used is
>>> Java 8 or a previous version. If it is Java 8, Lambda calculus can be used
>>> to make some things easier, especially while working with spark.
>>> Thanks and Regards,
>>> Anurag Kashyap
>>>
>>>
>>> On Mon, Mar 7, 2016 at 8:05 PM, Supun Sethunga  wrote:
>>>
 [Adding WSO2 Developers mailing list]

 Hi Anurag,

 Please find my comments inline.

 First, which boosting model are we going to target-Ada boosting or some
> other model, or various models?

 You can pick any method as you wish.


 Second, what method would you suggest I use to combine the results of
> the models in stacking- is it going to be a single layer LR model?

 This has to be a user input (i.e. user should be able to pick any
 algorithm for the base-learners as well as for the meta-learner).

 Regards,
 Supun

 On Mon, Mar 7, 2016 at 5:15 PM, Anurag Kashyap 
 wrote:

> Hi Supun,
>  I've begun reading the source rom link 3. I had a couple of
> questions. First, which boosting model are we going to target-Ada boosting
> or some other model, or various models? Second, what method would you
> suggest I use to combine the results of the models in stacking- is it 
> going
> to be a single layer LR model?
>
>
>
> On Sun, Mar 6, 2016 at 6:25 PM, Supun Sethunga 
> wrote:
>
>> Hi Anurag,
>>
>> Thank you for your interest in the above project. You can start by
>> getting to know the Machine Learner product by downloading it and running
>> it (Please use link [1] to download). Official documentation [2] will 
>> help
>> you on using the product. You can find the source code base at [3] and 
>> [4].
>> As we have mentioned in the project proposal as well, the main
>> objective is to integrate ensemble support for the existing flow of the
>> WSO2 Machine Learner. We are focusing on the three methods: Bagging,
>> Boosting and Stacking.
>>
>> Please feel free to raise, if you have any questions or any unclear
>> points.
>>
>> [1] http://wso2.com/products/machine-learner/
>> [2] https://docs.wso2.com/display/ML100/Introducing+Machine+Learner
>> [3] https://github.com/wso2/carbon-ml
>> [4] https://github.com/wso2/product-ml
>>
>> Regards,
>> Supun
>>
>> On Sun, Mar 6, 2016 at 12:08 PM, Anurag Kashyap > > wrote:
>>
>>> Hi Supun,
>>>  I'm currently working as a research assistant at NTU, Singapore for
>>> my bachelor thesis in my final year, I looked at the list of GSoC
>>> projects for WSO2 and I'm particularly interested in the project 
>>> *Ensemble
>>> 

Re: [Dev] GSoC project for building an ensemble of models

[Supun Sethunga]

Hi Anurag,

Yes, the master branch is using java 8, hence you can use lambda
expressions.

Regards,
Supun

On Mon, Mar 14, 2016 at 3:24 PM, Anurag Kashyap 
wrote:

> Hi Supun,
> One more question. I couldn't find if the version of Java being used is
> Java 8 or a previous version. If it is Java 8, Lambda calculus can be used
> to make some things easier, especially while working with spark.
> Thanks and Regards,
> Anurag Kashyap
>
>
> On Mon, Mar 7, 2016 at 8:05 PM, Supun Sethunga  wrote:
>
>> [Adding WSO2 Developers mailing list]
>>
>> Hi Anurag,
>>
>> Please find my comments inline.
>>
>> First, which boosting model are we going to target-Ada boosting or some
>>> other model, or various models?
>>
>> You can pick any method as you wish.
>>
>>
>> Second, what method would you suggest I use to combine the results of the
>>> models in stacking- is it going to be a single layer LR model?
>>
>> This has to be a user input (i.e. user should be able to pick any
>> algorithm for the base-learners as well as for the meta-learner).
>>
>> Regards,
>> Supun
>>
>> On Mon, Mar 7, 2016 at 5:15 PM, Anurag Kashyap 
>> wrote:
>>
>>> Hi Supun,
>>>  I've begun reading the source rom link 3. I had a couple of questions.
>>> First, which boosting model are we going to target-Ada boosting or some
>>> other model, or various models? Second, what method would you suggest I use
>>> to combine the results of the models in stacking- is it going to be a
>>> single layer LR model?
>>>
>>>
>>>
>>> On Sun, Mar 6, 2016 at 6:25 PM, Supun Sethunga  wrote:
>>>
 Hi Anurag,

 Thank you for your interest in the above project. You can start by
 getting to know the Machine Learner product by downloading it and running
 it (Please use link [1] to download). Official documentation [2] will help
 you on using the product. You can find the source code base at [3] and [4].
 As we have mentioned in the project proposal as well, the main
 objective is to integrate ensemble support for the existing flow of the
 WSO2 Machine Learner. We are focusing on the three methods: Bagging,
 Boosting and Stacking.

 Please feel free to raise, if you have any questions or any unclear
 points.

 [1] http://wso2.com/products/machine-learner/
 [2] https://docs.wso2.com/display/ML100/Introducing+Machine+Learner
 [3] https://github.com/wso2/carbon-ml
 [4] https://github.com/wso2/product-ml

 Regards,
 Supun

 On Sun, Mar 6, 2016 at 12:08 PM, Anurag Kashyap 
 wrote:

> Hi Supun,
>  I'm currently working as a research assistant at NTU, Singapore for
> my bachelor thesis in my final year, I looked at the list of GSoC
> projects for WSO2 and I'm particularly interested in the project *Ensemble
> Methods Support for WSO2 Machine Learner.* I am currently working on
> Ensemble methods for multi-label classification, so I have some experience
> in this field and I've read several scholarly articles about their
> implementation and benefit. I believe I can contribute to this project as 
> a
> GSoC intern. Eagerly awaiting a response!
>
> Yours Sincerely,
> *Anurag Kashyap*
> *Research Assistant*
> *Robotics-|| *
> *Nanyang Technological University, Singapore*
>
> [image: Blog] 
>



 --
 *Supun Sethunga*
 Software Engineer
 WSO2, Inc.
 http://wso2.com/
 lean | enterprise | middleware
 Mobile : +94 716546324

>>>
>>>
>>>
>>> --
>>> Yours sincerely,
>>>
>>> *Anurag Kashyap*
>>> *MSc.(Tech.)* Information Systems,
>>> *BITS Pilani K K Birla Goa Campus*
>>>
>>> [image: Blog] 
>>>
>>
>>
>>
>> --
>> *Supun Sethunga*
>> Software Engineer
>> WSO2, Inc.
>> http://wso2.com/
>> lean | enterprise | middleware
>> Mobile : +94 716546324
>>
>
>
>
> --
> Yours sincerely,
>
> *Anurag Kashyap*
> *MSc.(Tech.)* Information Systems,
> *BITS Pilani K K Birla Goa Campus*
>
> [image: Blog] 
>



-- 
*Supun Sethunga*
Software Engineer
WSO2, Inc.
http://wso2.com/
lean | enterprise | middleware
Mobile : +94 716546324
___
Dev mailing list
Dev@wso2.org
http://wso2.org/cgi-bin/mailman/listinfo/dev

Re: [Dev] GSoC project for building an ensemble of models

[Supun Sethunga]

[Adding WSO2 Developers mailing list]

Hi Anurag,

Please find my comments inline.

First, which boosting model are we going to target-Ada boosting or some
> other model, or various models?

You can pick any method as you wish.


Second, what method would you suggest I use to combine the results of the
> models in stacking- is it going to be a single layer LR model?

This has to be a user input (i.e. user should be able to pick any algorithm
for the base-learners as well as for the meta-learner).

Regards,
Supun

On Mon, Mar 7, 2016 at 5:15 PM, Anurag Kashyap  wrote:

> Hi Supun,
>  I've begun reading the source rom link 3. I had a couple of questions.
> First, which boosting model are we going to target-Ada boosting or some
> other model, or various models? Second, what method would you suggest I use
> to combine the results of the models in stacking- is it going to be a
> single layer LR model?
>
>
>
> On Sun, Mar 6, 2016 at 6:25 PM, Supun Sethunga  wrote:
>
>> Hi Anurag,
>>
>> Thank you for your interest in the above project. You can start by
>> getting to know the Machine Learner product by downloading it and running
>> it (Please use link [1] to download). Official documentation [2] will help
>> you on using the product. You can find the source code base at [3] and [4].
>> As we have mentioned in the project proposal as well, the main objective
>> is to integrate ensemble support for the existing flow of the WSO2 Machine
>> Learner. We are focusing on the three methods: Bagging, Boosting and
>> Stacking.
>>
>> Please feel free to raise, if you have any questions or any unclear
>> points.
>>
>> [1] http://wso2.com/products/machine-learner/
>> [2] https://docs.wso2.com/display/ML100/Introducing+Machine+Learner
>> [3] https://github.com/wso2/carbon-ml
>> [4] https://github.com/wso2/product-ml
>>
>> Regards,
>> Supun
>>
>> On Sun, Mar 6, 2016 at 12:08 PM, Anurag Kashyap 
>> wrote:
>>
>>> Hi Supun,
>>>  I'm currently working as a research assistant at NTU, Singapore for my
>>> bachelor thesis in my final year, I looked at the list of GSoC projects
>>> for WSO2 and I'm particularly interested in the project *Ensemble
>>> Methods Support for WSO2 Machine Learner.* I am currently working on
>>> Ensemble methods for multi-label classification, so I have some experience
>>> in this field and I've read several scholarly articles about their
>>> implementation and benefit. I believe I can contribute to this project as a
>>> GSoC intern. Eagerly awaiting a response!
>>>
>>> Yours Sincerely,
>>> *Anurag Kashyap*
>>> *Research Assistant*
>>> *Robotics-|| *
>>> *Nanyang Technological University, Singapore*
>>>
>>> [image: Blog] 
>>>
>>
>>
>>
>> --
>> *Supun Sethunga*
>> Software Engineer
>> WSO2, Inc.
>> http://wso2.com/
>> lean | enterprise | middleware
>> Mobile : +94 716546324
>>
>
>
>
> --
> Yours sincerely,
>
> *Anurag Kashyap*
> *MSc.(Tech.)* Information Systems,
> *BITS Pilani K K Birla Goa Campus*
>
> [image: Blog] 
>



-- 
*Supun Sethunga*
Software Engineer
WSO2, Inc.
http://wso2.com/
lean | enterprise | middleware
Mobile : +94 716546324
___
Dev mailing list
Dev@wso2.org
http://wso2.org/cgi-bin/mailman/listinfo/dev

Re: [Dev] GSoC project for building an ensemble of models

[Supun Sethunga]

Hi Anurag,

Thank you for your interest in the above project. You can start by getting
to know the Machine Learner product by downloading it and running it
(Please use link [1] to download). Official documentation [2] will help you
on using the product. You can find the source code base at [3] and [4].
As we have mentioned in the project proposal as well, the main objective is
to integrate ensemble support for the existing flow of the WSO2 Machine
Learner. We are focusing on the three methods: Bagging, Boosting and
Stacking.

Please feel free to raise, if you have any questions or any unclear points.

[1] http://wso2.com/products/machine-learner/
[2] https://docs.wso2.com/display/ML100/Introducing+Machine+Learner
[3] https://github.com/wso2/carbon-ml
[4] https://github.com/wso2/product-ml

Regards,
Supun

On Sun, Mar 6, 2016 at 12:08 PM, Anurag Kashyap 
wrote:

> Hi Supun,
>  I'm currently working as a research assistant at NTU, Singapore for my
> bachelor thesis in my final year, I looked at the list of GSoC projects
> for WSO2 and I'm particularly interested in the project *Ensemble Methods
> Support for WSO2 Machine Learner.* I am currently working on Ensemble
> methods for multi-label classification, so I have some experience in this
> field and I've read several scholarly articles about their implementation
> and benefit. I believe I can contribute to this project as a GSoC intern.
> Eagerly awaiting a response!
>
> Yours Sincerely,
> *Anurag Kashyap*
> *Research Assistant*
> *Robotics-|| *
> *Nanyang Technological University, Singapore*
>
> [image: Blog] 
>



-- 
*Supun Sethunga*
Software Engineer
WSO2, Inc.
http://wso2.com/
lean | enterprise | middleware
Mobile : +94 716546324
___
Dev mailing list
Dev@wso2.org
http://wso2.org/cgi-bin/mailman/listinfo/dev