A menagerie of machine learning utilities.
Currently implements k_fold learning and k_fold comparative performance plots via via Real.
It is likely that bootstrapping will be added soon and also a couple of additional types of comparative plots.
In SWI It is highly recommended that you install from within by using its package manager.
To install simply do:
?- pack_install(mlu).
And then load by:
?- [library(mlu)].
The example runs 10-fold cross validation on 2 trivial learners and
generates some random-based prediction results that are then displayed
as jitter plots via Real.
module documentation:
mlu.html
sources: packs/mlu
This library depends on packs:
lib, os_lib,
debug_call, and
pack_errors.
It also benefits from the presence of pack
Real.
In the case of pack(lib), this is picked by the pack manager interactively
during installation.
Whereas packs: options and pack_errors are interactively installed during
first load.
In all cases the user is asked before downloading and installing anything.
If you want support from packs Real and mtx install them manually via the pack manager.
Nicos Angelopoulos
---
London
March, 2017