Note: this page autoupdates while a run is in progress
(see end of log file)
===== MAIN: learn based on training data =====
=== START program1: ./run learn ../dataset3/train
=== START program2: ./run learn ../program1/data
Using hyperparameter c = 0.1
Reading training examples... (8140 examples) done
Training set properties: 12 features, 2 classes
Iter 1: .........*(NumConst=1, SV=1, CEps=100.0000, QPEps=0.0000)
Iter 2: *(NumConst=2, SV=1, CEps=79.3278, QPEps=8.9247)
Iter 3: *(NumConst=3, SV=2, CEps=10.3415, QPEps=5.0443)
Iter 4: .........*(NumConst=4, SV=2, CEps=5.7129, QPEps=2.5143)
Iter 5: *(NumConst=5, SV=2, CEps=0.9454, QPEps=0.2722)
Iter 6: .........*(NumConst=6, SV=2, CEps=0.4157, QPEps=0.1660)
Iter 7: *(NumConst=7, SV=2, CEps=0.2060, QPEps=0.0071)
Iter 8: .........(NumConst=7, SV=2, CEps=0.0533, QPEps=0.0071)
Final epsilon on KKT-Conditions: 0.05329
Upper bound on duality gap: 0.00584
Dual objective value: dval=8.90984
Primal objective value: pval=8.91568
Total number of constraints in final working set: 7 (of 7)
Number of iterations: 8
Number of calls to 'find_most_violated_constraint': 32560
Number of SV: 2
Norm of weight vector: |w|=0.90383
Value of slack variable (on working set): xi=85.01893
Value of slack variable (global): xi=85.07223
Norm of longest difference vector: ||Psi(x,y)-Psi(x,ybar)||=42.67137
Runtime in cpu-seconds: 0.25
Final number of constraints in cache: 16280
Compacting linear model...done
Writing learned model...done
=== END program2: ./run learn ../program1/data --- OK [0s]
=== END program1: ./run learn ../dataset3/train --- OK [1s]
===== MAIN: predict/evaluate on train data =====
=== START program4: ./run stripLabels ../dataset3/train ../program0/evalTrain.in
=== END program4: ./run stripLabels ../dataset3/train ../program0/evalTrain.in --- OK [1s]
=== START program1: ./run predict ../program0/evalTrain.in ../program0/evalTrain.out
=== START program2: ./run predict ../program0/evalTrain.in ../program0/evalTrain.out.multiclass-output
Reading test examples...
ERROR: The class label '0.000000' of example number 1 is not greater than '1'!
./run:22: Failed (RuntimeError)
=== END program2: ./run predict ../program0/evalTrain.in ../program0/evalTrain.out.multiclass-output --- FAILED [0s]
=== END program1: ./run predict ../program0/evalTrain.in ../program0/evalTrain.out --- FAILED [0s]
supervised-learning: Main entry for supervised learning for training and testing a program on a dataset.
(learner:Program) binary-to-multi: Allows multiclass classification to be run on binary classification datasets (trivial reduction).
(multiclassLearner:Program[MulticlassClassification]) svmlight_multiclass: SVMlight for multiclass classification (http://svmlight.joachims.org/svm_multiclass.html)
Go to the page for the run and look at the log file for signs of the responsible error.
You can also download the run and run it locally on your machine (a README file should
be included in the download which provides more information).
We said that a run was simply a program/dataset pair, but that's not the full story.
A run actually includes other helper programs such as the evaluation program and
various programs for reductions (e.g., one-versus-all, hyperparameter tuning).
More formally, a run is a given by a run specification,
which can be found on the page for any run.
A run specification is a tree where each internal node represents a program
and its children represents the arguments to be passed into its constructor.
For example, the one-versus-all program takes your binary classification program
as a constructor argument and behaves like a multiclass classification program.