Package: innsight 0.3.0

Niklas Koenen

innsight: Get the Insights of Your Neural Network

Interpretation methods for analyzing the behavior and individual predictions of modern neural networks in a three-step procedure: Converting the model, running the interpretation method, and visualizing the results. Implemented methods are, e.g., 'Connection Weights' described by Olden et al. (2004) <doi:10.1016/j.ecolmodel.2004.03.013>, layer-wise relevance propagation ('LRP') described by Bach et al. (2015) <doi:10.1371/journal.pone.0130140>, deep learning important features ('DeepLIFT') described by Shrikumar et al. (2017) <arxiv:1704.02685> and gradient-based methods like 'SmoothGrad' described by Smilkov et al. (2017) <arxiv:1706.03825>, 'Gradient x Input' described by Baehrens et al. (2009) <arxiv:0912.1128> or 'Vanilla Gradient'.

Authors:Niklas Koenen [aut, cre], Raphael Baudeu [ctb]

innsight_0.3.0.tar.gz
innsight_0.3.0.zip(r-4.5)innsight_0.3.0.zip(r-4.4)innsight_0.3.0.zip(r-4.3)
innsight_0.3.0.tgz(r-4.4-any)innsight_0.3.0.tgz(r-4.3-any)
innsight_0.3.0.tar.gz(r-4.5-noble)innsight_0.3.0.tar.gz(r-4.4-noble)
innsight_0.3.0.tgz(r-4.4-emscripten)innsight_0.3.0.tgz(r-4.3-emscripten)
innsight.pdf |innsight.html✨
innsight/json (API)
NEWS

Peer review:

Bug tracker:https://github.com/bips-hb/innsight/issues

On CRAN:

28 exports 26 stars 2.46 score 42 dependencies 29 scripts 1.1k downloads

Last updated 9 months agofrom:6dea46870c. Checks:OK: 5 NOTE: 2. Indexed: yes.

Exports:AgnosticWrapperConnectionWeightsconvertConverterDeepLiftDeepSHAPExpectedGradientget_resultGradientIntegratedGradientLIMELRPplotplot_globalprintrun_cwrun_deepliftrun_deepshaprun_expgradrun_gradrun_intgradrun_limerun_lrprun_shaprun_smoothgradSHAPshowSmoothGrad

Dependencies:backportsbitbit64callrcheckmateclicolorspacecorodescellipsisfansifarverggplot2gluegtableisobandjsonlitelabelinglatticelifecyclemagrittrMASSMatrixmgcvmunsellnlmepillarpkgconfigprocessxpsR6RColorBrewerRcpprlangsafetensorsscalestibbletorchutf8vctrsviridisLitewithr