# Your wekaExperiment example report

## Experiment: weka

• Methods: J48, NaiveBayes, OneR, RandomForest
• Problems: anneal, audiology, balance-scale, car, glass, horsecolic, hypothyroid, ionosphere, liver-disorders, lymph, primary-tumor, soybean, vehicle, vote, vowel
• Parameters:
• featureSelection [no,yes] (instantiated)
• Outputs: accuracy, trainingTime
##### History:
1. Experiment weka loaded from a data set.
2. Parameters 'fold' have been removed using the function 'mean'
3. A set of pairs 'column:[defaultValues]' has been used to subset the experiment: featureSelection:[no]
4. Methods has been instanciated with the parameters: featureSelection

## Experiment: weka

• Methods: NaiveBayes, J48, RandomForest, OneR
• Problems: hypothyroid, liver-disorders, vehicle, car, ionosphere, soybean, vote, anneal, lymph, vowel, audiology, glass, primary-tumor, balance-scale, horsecolic
• Parameters:
• featureSelection [no,yes]
• fold [0,1,2,3,4,5,6,7,8,9]
• Outputs: accuracy, trainingTime
##### History:
1. Experiment weka loaded from a data set.

## Experiment: weka

• Methods: J48, NaiveBayes, OneR, RandomForest
• Problems: anneal, audiology, balance-scale, car, glass, horsecolic, hypothyroid, ionosphere, liver-disorders, lymph, primary-tumor, soybean, vehicle, vote, vowel
• Parameters:
• nuevoPar [nuevoVal]
• featureSelection [no,yes] (instantiated)
• Outputs: accuracy, trainingTime
##### History:
1. Experiment weka loaded from a data set.
2. Parameters 'fold' have been removed using the function 'mean'
3. A set of pairs 'column:[defaultValues]' has been used to subset the experiment: featureSelection:[no]
4. Methods has been instanciated with the parameters: featureSelection
5. New parameters have been added with default values: nuevoPar:nuevoVal

## Friedman Test for output "accuracy"

From experiment: weka

Friedman test for output accuracy
• H0: all methods being equivalent.
• H1: at least one method is not equivalent.
• Strategy for ranks: maximize
• Computed Chi squared statistic with 3 degrees of freedom: 20.84
• Computed p-value: 0.000113645596289616
• Outcome for α = 0.05: H0 Rejected

## Results for output "accuracy"

From experiment: weka

## Cumulative Ranking for Var accuracy

From experiment: weka

## Summary of Control post-hoc test for output "accuracy"

From experiment: weka
• Test: Control post-hoc test
• Adjust Method: Holm
• α = 0.05
• Boldfaced results represent non-rejected hipotheses

methodpvaluerankwintieloss
RandomForest-1.67---
J485.7161e-011.93906
NaiveBayes3.2419e-022.801104
OneR1.2329e-043.601500


\begin{tabular}{lrrrrr}
\hline
method & pvalue & rank & win & tie & loss\\
\hline
RandomForest & - & 1.67 & - & - & -\\
J48 & \bf 5.7161e-01 & 1.93 & 9 & 0 & 6\\
NaiveBayes & 3.2419e-02 & 2.80 & 11 & 0 & 4\\
OneR & 1.2329e-04 & 3.60 & 15 & 0 & 0\\
\hline
\end{tabular}


## Distribution of ranks for output "accuracy"

From experiment: weka

## Detailed results for output(s) "accuracy"

From experiment: weka

methodannealaudiologybalance-scalecarglasshorsecolichypothyroidionosphereliver-disorderslymphprimary-tumorsoybeanvehiclevotevowel
J4898.775375.296475.714377.886166.883185.322899.416787.476261.815177.047640.704182.402073.155596.997952.2222
NaiveBayes86.528168.656187.846973.135244.502278.235795.414279.468249.302582.476250.730888.416544.431489.656459.6970
OneR83.633046.561336.354370.008153.831281.516596.341578.341353.042074.285725.329824.912652.605095.618426.3636
RandomForest99.107471.778679.703181.878077.619184.211799.310693.746064.092479.142840.695284.750673.043496.538064.6465


\begin{tabular}{lrrrrrrrrrrrrrrr}
\hline
method & anneal & audiology & balance-scale & car & glass & horsecolic & hypothyroid & ionosphere & liver-disorders & lymph & primary-tumor & soybean & vehicle & vote & vowel\\
\hline
J48 & 98.7753 & \bf 75.2964 & 75.7143 & 77.8861 & 66.8831 & \bf 85.3228 & \bf 99.4167 & 87.4762 & 61.8151 & 77.0476 & 40.7041 & 82.4020 & \bf 73.1555 & \bf 96.9979 & 52.2222\\
NaiveBayes & 86.5281 & 68.6561 & \bf 87.8469 & 73.1352 & 44.5022 & 78.2357 & 95.4142 & 79.4682 & 49.3025 & \bf 82.4762 & \bf 50.7308 & \bf 88.4165 & 44.4314 & 89.6564 & 59.6970\\
OneR & 83.6330 & 46.5613 & 36.3543 & 70.0081 & 53.8312 & 81.5165 & 96.3415 & 78.3413 & 53.0420 & 74.2857 & 25.3298 & 24.9126 & 52.6050 & 95.6184 & 26.3636\\
RandomForest & \bf 99.1074 & 71.7786 & 79.7031 & \bf 81.8780 & \bf 77.6191 & 84.2117 & 99.3106 & \bf 93.7460 & \bf 64.0924 & 79.1428 & 40.6952 & 84.7506 & 73.0434 & 96.5380 & \bf 64.6465\\
\hline
\end{tabular}


##### exreport, Copyright 2015 by Jacinto Arias and Javier Cozar
Generated using R version 3.2.1 (2015-06-18)