Discussion:
Proposal Submission - GSoC 2017
Aditya
2017-04-02 23:07:29 UTC
Permalink
Hello everyone,

I've completed making my proposal and have submitted it successfully in the
Google Summer of Code webpage.

It would be great if some of you could go through it and provide feedback.
Also, is there a specific template that should be followed for applying to
ASF?

This
<https://docs.google.com/document/d/1YSHxDk7H_ybYxDEEy9fUUCfpI3KEFVCCH2l1gmKra2Q/edit?usp=sharing>
is
the link to my proposal.


Thanks,
Aditya
Trevor Grant
2017-04-04 18:26:59 UTC
Permalink
This looks really good.

It's also ambitious (which can be good). Want to make sure you think you
can achieve what you set out to.

I would also say tests on the naive DBSCAN should come before Implementing
the R-Tree approach.

You said seperate 'module" this may just be semantics, but I would think
these things would live in
org.apache.mahout.math.algorithms.classifiers.dbscan
(i.e. not a separate module like org.apache.mahout.dbscan).

Would love to hear any other though, esp if anyone is more familiar with
this algo.

tg


Trevor Grant
Data Scientist
https://github.com/rawkintrevo
http://stackexchange.com/users/3002022/rawkintrevo
http://trevorgrant.org

*"Fortunate is he, who is able to know the causes of things." -Virgil*
Post by Aditya
Hello everyone,
I've completed making my proposal and have submitted it successfully in the
Google Summer of Code webpage.
It would be great if some of you could go through it and provide feedback.
Also, is there a specific template that should be followed for applying to
ASF?
This
<https://docs.google.com/document/d/1YSHxDk7H_
ybYxDEEy9fUUCfpI3KEFVCCH2l1gmKra2Q/edit?usp=sharing>
is
the link to my proposal.
Thanks,
Aditya
Morris Riedel
2017-04-04 19:41:24 UTC
Permalink
Dear Aditya,

I'm one of the co-authors of this publication from one of my PhD students:

[1] Markus Götz, Christian Bodenstein, and Morris Riedel. 2015. HPDBSCAN: highly parallel DBSCAN. In Proceedings of the Workshop on Machine Learning in High-Performance Computing Environments (MLHPC '15). ACM, New York, NY, USA, , Article 2 , 10 pages. DOI=http://dx.doi.org/10.1145/2834892.2834894

We are highly interested in your GsoC proposal and offer
you support - if needed just get in contact to us and Markus.

Implementing this in Mahout is a great idea.

We use other elements like collaborative filtering in Mahout, so not only HPC, etc.

But DBSCAN here would be great to see in Mahout!

Fingers crossed!

Take care,
Morris

-----Ursprüngliche Nachricht-----
Von: Trevor Grant [mailto:***@gmail.com]
Gesendet: Tuesday, April 04, 2017 6:27 PM
An: ***@mahout.apache.org
Betreff: Re: Proposal Submission - GSoC 2017

This looks really good.

It's also ambitious (which can be good). Want to make sure you think you
can achieve what you set out to.

I would also say tests on the naive DBSCAN should come before Implementing
the R-Tree approach.

You said seperate 'module" this may just be semantics, but I would think
these things would live in
org.apache.mahout.math.algorithms.classifiers.dbscan
(i.e. not a separate module like org.apache.mahout.dbscan).

Would love to hear any other though, esp if anyone is more familiar with
this algo.

tg


Trevor Grant
Data Scientist
https://github.com/rawkintrevo
http://stackexchange.com/users/3002022/rawkintrevo
http://trevorgrant.org

*"Fortunate is he, who is able to know the causes of things." -Virgil*
Post by Aditya
Hello everyone,
I've completed making my proposal and have submitted it successfully in the
Google Summer of Code webpage.
It would be great if some of you could go through it and provide feedback.
Also, is there a specific template that should be followed for applying to
ASF?
This
<https://docs.google.com/document/d/1YSHxDk7H_
ybYxDEEy9fUUCfpI3KEFVCCH2l1gmKra2Q/edit?usp=sharing>
is
the link to my proposal.
Thanks,
Aditya
------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------
Forschungszentrum Juelich GmbH
52425 Juelich
Sitz der Gesellschaft: Juelich
Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher
Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
Prof. Dr. Sebastian M. Schmidt
------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------
Aditya
2017-04-06 22:00:30 UTC
Permalink
Dear Dr. Riedel,

It is indeed a pleasure to hear from you. HPDBSCAN was one of the first
algorithms that I studied when I started off my work on a distributed
algorithm.

It is very kind of you to offer me support. I hope the proposal gets
accepted and will get in touch once I'm working on the implementation..

Regards,
Aditya
Post by Morris Riedel
Dear Aditya,
highly parallel DBSCAN. In Proceedings of the Workshop on Machine Learning
in High-Performance Computing Environments (MLHPC '15). ACM, New York, NY,
USA, , Article 2 , 10 pages. DOI=http://dx.doi.org/10.1145/2834892.2834894
We are highly interested in your GsoC proposal and offer
you support - if needed just get in contact to us and Markus.
Implementing this in Mahout is a great idea.
We use other elements like collaborative filtering in Mahout, so not only HPC, etc.
But DBSCAN here would be great to see in Mahout!
Fingers crossed!
Take care,
Morris
-----UrsprÃŒngliche Nachricht-----
Gesendet: Tuesday, April 04, 2017 6:27 PM
Betreff: Re: Proposal Submission - GSoC 2017
This looks really good.
It's also ambitious (which can be good). Want to make sure you think you
can achieve what you set out to.
I would also say tests on the naive DBSCAN should come before Implementing
the R-Tree approach.
You said seperate 'module" this may just be semantics, but I would think
these things would live in
org.apache.mahout.math.algorithms.classifiers.dbscan
(i.e. not a separate module like org.apache.mahout.dbscan).
Would love to hear any other though, esp if anyone is more familiar with
this algo.
tg
Trevor Grant
Data Scientist
https://github.com/rawkintrevo
http://stackexchange.com/users/3002022/rawkintrevo
http://trevorgrant.org
*"Fortunate is he, who is able to know the causes of things." -Virgil*
Post by Aditya
Hello everyone,
I've completed making my proposal and have submitted it successfully in
the
Post by Aditya
Google Summer of Code webpage.
It would be great if some of you could go through it and provide
feedback.
Post by Aditya
Also, is there a specific template that should be followed for applying
to
Post by Aditya
ASF?
This
<https://docs.google.com/document/d/1YSHxDk7H_
ybYxDEEy9fUUCfpI3KEFVCCH2l1gmKra2Q/edit?usp=sharing>
is
the link to my proposal.
Thanks,
Aditya
------------------------------------------------------------
------------------------------------
------------------------------------------------------------
------------------------------------
Forschungszentrum Juelich GmbH
52425 Juelich
Sitz der Gesellschaft: Juelich
Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher
Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
Prof. Dr. Sebastian M. Schmidt
------------------------------------------------------------
------------------------------------
------------------------------------------------------------
------------------------------------
Morris Riedel
2017-04-06 22:09:31 UTC
Permalink
Dear Aditya,

With pleasure, good to see Mahout evolving!


Let us know once we know about GSoc 2017, otherwise
We might another channel to fund like EU funding, or
national funding streams....

Take care,
Morris


-----Ursprüngliche Nachricht-----
Von: Aditya [mailto:***@gmail.com]
Gesendet: Friday, April 07, 2017 12:01 AM
An: ***@mahout.apache.org
Betreff: Re: AW: Proposal Submission - GSoC 2017

Dear Dr. Riedel,

It is indeed a pleasure to hear from you. HPDBSCAN was one of the first
algorithms that I studied when I started off my work on a distributed
algorithm.

It is very kind of you to offer me support. I hope the proposal gets
accepted and will get in touch once I'm working on the implementation..

Regards,
Aditya
Post by Morris Riedel
Dear Aditya,
highly parallel DBSCAN. In Proceedings of the Workshop on Machine Learning
in High-Performance Computing Environments (MLHPC '15). ACM, New York, NY,
USA, , Article 2 , 10 pages. DOI=http://dx.doi.org/10.1145/2834892.2834894
We are highly interested in your GsoC proposal and offer
you support - if needed just get in contact to us and Markus.
Implementing this in Mahout is a great idea.
We use other elements like collaborative filtering in Mahout, so not only HPC, etc.
But DBSCAN here would be great to see in Mahout!
Fingers crossed!
Take care,
Morris
-----Ursprüngliche Nachricht-----
Gesendet: Tuesday, April 04, 2017 6:27 PM
Betreff: Re: Proposal Submission - GSoC 2017
This looks really good.
It's also ambitious (which can be good). Want to make sure you think you
can achieve what you set out to.
I would also say tests on the naive DBSCAN should come before Implementing
the R-Tree approach.
You said seperate 'module" this may just be semantics, but I would think
these things would live in
org.apache.mahout.math.algorithms.classifiers.dbscan
(i.e. not a separate module like org.apache.mahout.dbscan).
Would love to hear any other though, esp if anyone is more familiar with
this algo.
tg
Trevor Grant
Data Scientist
https://github.com/rawkintrevo
http://stackexchange.com/users/3002022/rawkintrevo
http://trevorgrant.org
*"Fortunate is he, who is able to know the causes of things." -Virgil*
Post by Aditya
Hello everyone,
I've completed making my proposal and have submitted it successfully in
the
Post by Aditya
Google Summer of Code webpage.
It would be great if some of you could go through it and provide
feedback.
Post by Aditya
Also, is there a specific template that should be followed for applying
to
Post by Aditya
ASF?
This
<https://docs.google.com/document/d/1YSHxDk7H_
ybYxDEEy9fUUCfpI3KEFVCCH2l1gmKra2Q/edit?usp=sharing>
is
the link to my proposal.
Thanks,
Aditya
------------------------------------------------------------
------------------------------------
------------------------------------------------------------
------------------------------------
Forschungszentrum Juelich GmbH
52425 Juelich
Sitz der Gesellschaft: Juelich
Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher
Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
Prof. Dr. Sebastian M. Schmidt
------------------------------------------------------------
------------------------------------
------------------------------------------------------------
------------------------------------
Loading...