OpenML
1665019

Run 1665019

Task 9952 (Supervised Classification) phoneme Uploaded 19-07-2016 by Jan van Rijn
0 likes downloaded by 0 people 0 issues 0 downvotes , 0 total downloads
Issue #Downvotes for this reason By


Flow

weka.SPegasos(6)S. Shalev-Shwartz, Y. Singer, N. Srebro: Pegasos: Primal Estimated sub-GrAdient SOlver for SVM. In: 24th International Conference on MachineLearning, 807-814, 2007.
weka.SPegasos(6)_F0
weka.SPegasos(6)_L1.0E-4
weka.SPegasos(6)_E500

Result files

xml
Description

XML file describing the run, including user-defined evaluation measures.

model
Model readable

A human-readable description of the model that was built.

model
Model serialized

A serialized description of the model that can be read by the tool that generated it.

arff
Predictions

ARFF file with instance-level predictions generated by the model.

17 Evaluation measures

Area under ROC curve
0.7598
Per class
Cross-validation details (10-fold Crossvalidation)
F measure
0.7615
Per class
Cross-validation details (10-fold Crossvalidation)
Kappa
0.465
Cross-validation details (10-fold Crossvalidation)
KB relative information score
1920.7814
Cross-validation details (10-fold Crossvalidation)
Mean absolute error
0.248
Cross-validation details (10-fold Crossvalidation)
Mean prior absolute error
0.4147
Cross-validation details (10-fold Crossvalidation)
Number of instances
5404
Per class
Cross-validation details (10-fold Crossvalidation)
Precision
0.7915
Per class
Cross-validation details (10-fold Crossvalidation)
Predictive accuracy
0.752
Cross-validation details (10-fold Crossvalidation)
Prior entropy
0.8733
Cross-validation details (10-fold Crossvalidation)
Recall
0.752
Per class
Cross-validation details (10-fold Crossvalidation)
Relative absolute error
0.5979
Cross-validation details (10-fold Crossvalidation)
Root mean prior squared error
0.4554
Cross-validation details (10-fold Crossvalidation)
Root mean squared error
0.498
Cross-validation details (10-fold Crossvalidation)
Root relative squared error
1.0936
Cross-validation details (10-fold Crossvalidation)