DALEX: Descriptive mAchine Learning EXplanations

Machine Learning (ML) models are widely used and have various applications in classification or regression. Models created with boosting, bagging, stacking or similar techniques are often used due to their high performance, but such black-box models usually lack of interpretability. 'DALEX' package contains various explainers that help to understand the link between input variables and model output. The single_variable() explainer extracts conditional response of a model as a function of a single selected variable. It is a wrapper over packages 'pdp' and 'ALEPlot'. The single_prediction() explainer attributes arts of model prediction to articular variables used in the model. It is a wrapper over 'breakDown' package. The variable_dropout() explainer assess variable importance based on consecutive permutations. All these explainers can be plotted with generic plot() function and compared across different models.

Version: 0.2.0
Depends: R (≥ 3.0)
Imports: pdp, ggplot2, ALEPlot, breakDown, factorMerger, ggpubr
Suggests: gbm, randomForest, xgboost
Published: 2018-04-05
Author: Przemyslaw Biecek [aut, cre]
Maintainer: Przemyslaw Biecek <przemyslaw.biecek at gmail.com>
BugReports: https://github.com/pbiecek/DALEX/issues
License: GPL-2 | GPL-3 [expanded from: GPL]
URL: https://pbiecek.github.io/DALEX/
NeedsCompilation: no
Materials: NEWS
CRAN checks: DALEX results


Reference manual: DALEX.pdf
Package source: DALEX_0.2.0.tar.gz
Windows binaries: r-devel: DALEX_0.2.0.zip, r-release: DALEX_0.2.0.zip, r-oldrel: DALEX_0.2.0.zip
OS X binaries: r-release: DALEX_0.2.0.tgz, r-oldrel: DALEX_0.2.0.tgz
Old sources: DALEX archive


Please use the canonical form https://CRAN.R-project.org/package=DALEX to link to this page.