[GitHub] flink pull request: [FLINK-2102] [ml] Add predict operation for La...

2015-06-02 Thread asfgit
Github user asfgit closed the pull request at:

https://github.com/apache/flink/pull/744


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---


[GitHub] flink pull request: [FLINK-2102] [ml] Add predict operation for La...

2015-06-02 Thread tillrohrmann
Github user tillrohrmann commented on the pull request:

https://github.com/apache/flink/pull/744#issuecomment-107897490
  
LGTM. I will merge it as a temporary solution for the manual evaluation.


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---


[GitHub] flink pull request: [FLINK-2102] [ml] Add predict operation for La...

2015-05-29 Thread thvasilo
Github user thvasilo commented on the pull request:

https://github.com/apache/flink/pull/744#issuecomment-106840541
  
@tillrohrmann I added a description for SVM and MLR


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---


[GitHub] flink pull request: [FLINK-2102] [ml] Add predict operation for La...

2015-05-29 Thread thvasilo
Github user thvasilo commented on the pull request:

https://github.com/apache/flink/pull/744#issuecomment-106832309
  
OK I'm on it.


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---


[GitHub] flink pull request: [FLINK-2102] [ml] Add predict operation for La...

2015-05-29 Thread tillrohrmann
Github user tillrohrmann commented on the pull request:

https://github.com/apache/flink/pull/744#issuecomment-106797811
  
Nice work @thvasilo. The only thing which is missing is the update to the 
online docs where we say which types are supported for which operation.


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---


[GitHub] flink pull request: [FLINK-2102] [ml] Add predict operation for La...

2015-05-29 Thread thvasilo
Github user thvasilo commented on the pull request:

https://github.com/apache/flink/pull/744#issuecomment-106745455
  
@sachingoel0101 Yes, that is partially covered by 
[FLINK-2108](https://issues.apache.org/jira/browse/FLINK-2108), and an 
evaluation framework (along with cross-validation) is in our immediate plans.


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---


[GitHub] flink pull request: [FLINK-2102] [ml] Add predict operation for La...

2015-05-28 Thread sachingoel0101
Github user sachingoel0101 commented on the pull request:

https://github.com/apache/flink/pull/744#issuecomment-106672515
  
One further possible addition would be to add a class to evaluate accuracy. 
We could provide statistics such as F1-score, Precision at k,etc. 
Where would that fit in the pipeline mechanism though? 


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---


[GitHub] flink pull request: [FLINK-2102] [ml] Add predict operation for La...

2015-05-28 Thread thvasilo
Github user thvasilo commented on the pull request:

https://github.com/apache/flink/pull/744#issuecomment-106488005
  
Currently there is a lot of boilerplate, but these should be minimized once 
the syntactic sugar for map with broadcast vars is merged.


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---


[GitHub] flink pull request: [FLINK-2102] [ml] Add predict operation for La...

2015-05-28 Thread thvasilo
GitHub user thvasilo opened a pull request:

https://github.com/apache/flink/pull/744

[FLINK-2102] [ml] Add predict operation for LabeledVector

Add predict function for labeled data for SVM and MLR.

These functions return for each example in the input DataSet[LabeledVector] 
a pair (Double, Double) - (truth, prediction)

This output can then be used to evaluate the performance of the Predictors

You can merge this pull request into a Git repository by running:

$ git pull https://github.com/thvasilo/flink predict-labeled

Alternatively you can review and apply these changes as the patch at:

https://github.com/apache/flink/pull/744.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

This closes #744


commit a57e464e704fae21b6ebe3c0afed3e21fb15b951
Author: Theodore Vasiloudis 
Date:   2015-05-28T16:51:17Z

Add predict function for labeled data for SVM and MLR.

These functions return for each example in the input DataSet[LabeledVector] 
a pair (truth, prediction)




---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---