search

h2oai / h2o-3

H2O is an Open Source, Distributed, Fast & Scalable Machine Learning Platform: Deep Learning, Gradient Boosting (GBM) & XGBoost, Random Forest, Generalized Linear Modeling (GLM with Elastic Net), K-Means, PCA, Generalized Additive Models (GAM), RuleFit, Support Vector Machine (SVM), Stacked Ensembles, Automatic Machine Learning (AutoML), etc.
http://h2o.ai
Apache License 2.0
6.9k stars 2k forks source link

Results of DL biased to 1 class of 2 output classes for balanced Data #11927

Open exalate-issue-sync[bot] opened 1 year ago

exalate-issue-sync[bot] commented 1 year ago

I am using H2o package on Python, to apply DL on a simple balanced Dataset of 2 classes "0" and "1". It is divided into 2 files of 80 % data used for training and 20% for validation purpose.

The results are always biased to class "1", regardless tuning parameters, due to the confusion metric.

Also, I notice the prediction results P0 and P1 are not biased to a certain class if I use threshold of 0.5, but the predict first column is using F1 measure.

How can I enhance the results of the confusion table? Is there a way to make the predict value to apply another threshold and hence changing all the metrics results related? Thanks in advance.

My code Snippet is as follows: model = H2ODeepLearningEstimator( activation="Tanh", hidden=[50,50,50], distribution="multinomial", score_interval=10, epochs=1000, input_dropout_ratio=0.2 ,adaptive_rate=True , rho=0.98, epsilon = 1e-7 ,seed=1 ,reproducible=True ) model.train( x=x, y=y, training_frame =train, validation_frame = test)

hasithjp commented 1 year ago

JIRA Issue Migration Info

Jira Issue: PUBDEV-5054 Assignee: New H2O Bugs Reporter: Sahar Ghanem State: Open Fix Version: N/A Attachments: N/A Development PRs: N/A