PSVM giving NaNs for some metrics

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.

Apache License 2.0

6.92k stars 2k forks source link

This issue was found in R with binary classification, but should also be checked in Python too:

{code:r}library(h2o) h2o.init() train <- h2o.importFile("https://s3.amazonaws.com/erin-data/higgs/higgs_train_10k.csv") test <- h2o.importFile("https://s3.amazonaws.com/erin-data/higgs/higgs_test_5k.csv") y <- "response" x <- setdiff(names(train), y)

train[,y] <- as.factor(train[,y]) test[,y] <- as.factor(test[,y]) psvm <- h2o.psvm(x = x, y = y, training_frame = train) h2o.performance(psvm, test){code}

Gives NaNs:

H2OBinomialMetrics: psvm MSE: 0.43 RMSE: 0.6557439 LogLoss: NaN Mean Per-Class Error: 0.4269508 AUC: NaN AUCPR: NaN Gini: NaN

h2oai / h2o-3

PSVM giving NaNs for some metrics #8323