[
https://issues.apache.org/jira/browse/MAHOUT-1856?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15833750#comment-15833750
]
ASF GitHub Bot commented on MAHOUT-1856:
----------------------------------------
Github user andrewpalumbo commented on the issue:
https://github.com/apache/mahout/pull/246
Trevor- looks really good to me- I've left some comments mainly about
hyperparameter being moved to fit(...) from model, I think that his makes sense
in many ways, E.g, When doing an highly iterative Hyperparameter search, It
would eliminate a good amount of overhead to call:
```aModel.fit(....,HyperParameters: Map["hParameter1" -> value ,
"hParameter2" -> ...])```
rather than re-constructing the entire class each time. As well as i
noted in line, I think that the `fit(...)` method should have the ability to
return at least a `List[double]` of errors per row if needed, So I would
suggest that it return `Any` rather than Unit in the base Traits. (unless the
convention that you're following is to rely on predict for this.
> Create a framework for new Mahout Clustering, Classification, and
> Optimization Algorithms
> ------------------------------------------------------------------------------------------
>
> Key: MAHOUT-1856
> URL: https://issues.apache.org/jira/browse/MAHOUT-1856
> Project: Mahout
> Issue Type: New Feature
> Affects Versions: 0.12.1
> Reporter: Andrew Palumbo
> Assignee: Trevor Grant
> Priority: Critical
> Fix For: 0.13.0
>
>
> To ensure that Mahout does not become "A loose bag of algorithms", Create
> basic traits with funtions common to each class of algorithm.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
