Skip to content

The current project introduces a script building a stacked learning ensemble, containing a single multilayered ANN (meta-learner) trained using the predictions of a number of ANNs (Level-0 learners).

License

Notifications You must be signed in to change notification settings

GeorgiosEtsias/Stacked-ANN-ensemble-MATLAB

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 
 
 

Repository files navigation

Stacked-ANN-ensemble

The current project introduces a script generating a stacked learning ensemble, containing a single multilayered ANN (Level-1 learner or meta-learner) trained using the predictions (meta-data) of a number of ANNs (Level-0 learners).

The Vinyl bromide dataset, that is available with the Deep Learning Toolbox™ was used for training and testing the ensemble. The dataset contains 16 input and 1 output variable and a total of 68308 observations.

Multiple script executions indicated that the stacked ensemble performed 20% better than the best individual ANN.

About

The current project introduces a script building a stacked learning ensemble, containing a single multilayered ANN (meta-learner) trained using the predictions of a number of ANNs (Level-0 learners).

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages