Models in qsaR


The package automatically build six different regression or classification models to evaluate the quality of the dataset.
In qsaR you can just type the command models(), to get all the things automatically in a easy fashion way.
The first result is showed in the figure bellow saved under the name “models-pls.png”.

models-pls
Figure 1. The pls model for the training and test series

The second result is showed in the figure bellow saved under the name “models-mars.png”.

models-mars
Figure 2. The mars model for the training and test series

The thirty result is showed in the figure bellow saved under the name “models-boosted_tree.png”.

models-boosted_tree
Figure 3. The Boosted Tree model for the training and test series


The fourth result is showed in the figure bellow saved under the name “models-random_forest.png”.

models-random_forest
Figure 4. The Random Forest model for the training and test series


The fifth result is showed in the figure bellow saved under the name “models-svm.png”.

models-svm
Figure 5. The Support Vector Machine model for the training and test series


The sixth result is showed in the figure bellow saved under the name “models-elastic_net.png”.

models-elastic_net
Figure 6. The Elastic Net model for the training and test series


Final analysis


An easy way to analyze all models in one place.

models-all


Final Observations


1- to compute all the models just type the command bias()
WARNING this could take some time, depends on the size of the matrix!
2 - For some reasons if lost some graph it’s possible to generate all of them just type plotmodels()
WARNING you need to compute the models before!