deep-significance
Enabling easy statistical significance testing for deep neural networks.
GPL-3.0 License
Bot releases are visible (Hide)
- Added feature to automatically select the maximum amount of jobs by using
num_jobs=-1
Published by Kaleidophon over 2 years ago
Changes:
- Updated documentation and code based on insights from paper experiments
- Speed and reliability improvements
- Deprecating some arguments, such as
num_samples,use_symmetryand addingnum_comparisons
Published by Kaleidophon almost 3 years ago
Add functions to help determine the right sample size, namely aso_uncertainty_reduction(), bootstrap_power_analysis()
Published by Kaleidophon almost 3 years ago
Fixing the bug described in #7 with commit a900e4a2d60278db211f9439a4f910e0aa53073b.
Published by Kaleidophon almost 3 years ago
Two critical hotfixes (see commit 23bf573). Additional minor fix for documentation in README.
Published by Kaleidophon about 3 years ago
Features:
- Make it easier to compare multiple model scores at once with
multi_aso(). - Results can be created as a
pandas.DataFramefor readability. - Both
aso()andmulti_aso()support aseedargument for replicability.
Published by Kaleidophon over 3 years ago
First release of deepsig! The package currently contains the following:
- Re-implementation of almost stochastic order (ASO) with optional multi-threading speed-up
- Extensive documentation and readme to guide practicioners
- Additional implementations of the permutation-randomization test, paired-bootstrap test and bonferroni correction (for p-values)
Published by Kaleidophon over 3 years ago
First release of deepsig! The package currently contains the following:
- Re-implementation of almost stochastic order (ASO) with optional multi-threading speed-up
- Extensive documentation and readme to guide practicioners
- Additional implementations of the permutation-randomization test, paired-bootstrap test and bonferroni correction (for p-values)
