Create learners document #5698
Create learners document #5698
Conversation
mairaw
added
waiting-on-reviews
and removed
waiting-on-reviews
waiting-on-feedback
|
@CESARDELATORRE will be helping with these docs too. He is now a collaborator on my forked docs repo so he should be able to take it from here. @GalOshri are you able to help review the table accuracy as a first step? |
| @@ -0,0 +1,67 @@ | |||
| # Learning Algorithms | |||
| The following table will help you understand the different learning algorithms (learners) included in ML.NET and the [machine learning tasks](tasks.md) that they can be used for. Click on the learner name for more in-depth explanations on each of these learning methods. | |||
There was a problem hiding this comment.
Perhaps "the machine learning tasks where they can be used"?
| | [Generalized Additive Models](#generalized-additive-models) |:heavy_check_mark:||:heavy_check_mark:|||| | ||
| | [K-Means++](#k-means) |||||:heavy_check_mark:|| | ||
| | [Linear SVM](#linear-svm) |:heavy_check_mark:|||||| | ||
| | [Logistic Regressor](#logistic-regressor) |:heavy_check_mark:|:heavy_check_mark:||||| |
There was a problem hiding this comment.
The name of the Logistic Regression learner has been updated (should just be Logistic Regression)
| ## Linear SVM | ||
|
|
||
|
|
||
| ## Logistic Regressor |
|
Looks good so far, thank you for creating this! |
| > [!NOTE] | ||
| > ML.NET is currently in Preview. Not all learning algorithms are currently supported. To submit a request for a certain learner, please open an issue in the [dotnet/machinelearning repository](https://github.com/dotnet/machinelearning/issues). | ||
|
|
||
| | Learner Name | Binary classification | Multi-class classification | Regression | Ranking | Clustering | Anomaly Detection | |
There was a problem hiding this comment.
In other docs we are spelling "Multiclass", not "Multi-class". I'd suggest to be consistent. in ML world "Multiclass" seems to be more common (https://en.wikipedia.org/wiki/Multiclass_classification, http://www.mit.edu/~9.520/spring09/Classes/multiclass.pdf, https://www.coursera.org/learn/machine-learning/lecture/gFpiW/multiclass-classification) .
| | [K-Means++](#k-means) |||||:heavy_check_mark:|| | ||
| | [Linear SVM](#linear-svm) |:heavy_check_mark:|||||| | ||
| | [Logistic Regressor](#logistic-regressor) |:heavy_check_mark:|:heavy_check_mark:||||| | ||
| | [Naïve Bayes](#naïve-bayes) ||:heavy_check_mark:||||| |
There was a problem hiding this comment.
Too many dots above the "i"? :) In our code and in majority of textbooks it's "Naive", which is also easier to search for users.
JRAlexander
modified the milestones:
Sprint 137 (06/11/18 - 06/29/18),
Sprint 138 (07/02/2018 - 07/20/2018)
JRAlexander
modified the milestones:
Sprint 138 (07/02/2018 - 07/20/2018),
Sprint 139 (7/23/2018 - 8/10/2018)
mairaw
modified the milestones:
Sprint 139 (7/23/2018 - 8/10/2018),
Sprint 141 (9/3/2018 - 9/21/2018)
JRAlexander
modified the milestones:
Sprint 141 (9/3/2018 - 9/21/2018),
Sprint 142 (9/24/2018 - 10/12/2018)
JRAlexander
modified the milestones:
Sprint 142 (9/24/2018 - 10/12/2018),
Sprint 143 (10/15/2018 - 11/2/2018)
Document to describe learning algorithms.
@GalOshri I will need help from your team to fill out the definitions/descriptions of the learners.
Internal Review URL