-
Notifications
You must be signed in to change notification settings - Fork 2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
PUBDEV-4940 implement UpliftRandomForest #5224
Conversation
5a27bf8
to
249c61a
Compare
f02cf61
to
df747d9
Compare
6df75fd
to
3ec1532
Compare
1efa255
to
cf8e9c5
Compare
2d03b1b
to
f003ea0
Compare
7065013
to
98a8667
Compare
98a8667
to
1affd8a
Compare
3f4a9e1
to
29292b0
Compare
b747f99
to
8df0e90
Compare
d83ee93
to
c43940b
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Great work @maurever! Looks good to me - great start - further refactoring will be needed but this is good enough to go into master
3d2bd01
to
3ae93d3
Compare
- added documentation pages - added params/algo to toctree; minor syntax updates, still need to add image to Uplift DRF; rst files only - added image; toctree algo shift - updated api comments - add comma - fix python example, it was failing on assertion error and uplift_model does not have plot_auuc() method - fix test in uplift_metric, it failing on assert - fix all available in
…g in scoring, improve API
… remove unused parameter
52186b7
to
5fd92a8
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM 👌 Thanks @maurever !
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Good job! 💯
JIRA:
https://h2oai.atlassian.net/jira/software/c/projects/PUBDEV/issues/PUBDEV-4940
Sources:
Existing implementation:
TODO:
nbins
parameter)auuc_type
to performance to get AUUC for unseen test dataIn this PR:
Improvements for future PR:
Implementation problems:
Uplift Random Forest API:
Uplift trees implementation is currently supported only for binomial classification.
API for Python:
H2OUpliftRandomForestEstimator
: train a modelperformance.auuc
: get the default AUUC from the performance objectmodel.auuc
: get the default AUUC from the modelperformance.auuc_table
: get all types of AUUCs from the performance objectmodel.auuc_table
: get all types of AUUCs from the modelplot_uplift
: plot uplift curve or returns values for x and y axis of the plot, you can set a metric typemake_metrics
: make H2OBinomialUpliftMetrics metricsAPI for R
h2o.upliftRandomForest
: train a modelh2o.performance
: returns H2OBinomialUpliftMetrics-
h2o.auuc
: get the default AUUCh2o.auuc_table
: get all types of AUUCsplot.H2OBinomialUpliftMetrics
: plot uplift curve or returns values for x and y-axis of the plot, you can set a metric typeh2o.make_metrics
: make H2OBinomialUpliftMetrics metricsPython Uplift curve:
R Uplift curve: