DALEX - moDel Agnostic Language for Exploration and eXplanation

Any unverified black box model is the path to failure. Opaqueness leads to distrust. Distrust leads to ignoration. Ignoration leads to rejection. DALEX package xrays any model and helps to explore and explain its behaviour. 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 direct interpretability. DALEX package contains various methods that help to understand the link between input variables and model output. Implemented methods help to explore the model on the level of a single instance as well as a level of the whole dataset. All model explainers are model agnostic and can be compared across different models. DALEX package is the cornerstone for 'DrWhy.AI' universe of packages for visual model exploration. Find more details in (Biecek 2018) <https://jmlr.org/papers/v19/18-416.html>.

Last updated

black-boxdalexdata-scienceexplainable-aiexplainable-artificial-intelligenceexplainable-mlexplanationsexplanatory-model-analysisfairnessimlinterpretabilityinterpretable-machine-learningmachine-learningmodel-visualizationpredictive-modelingresponsible-airesponsible-mlxai

13.53 score 1.5k stars 19 dependents 1.5k scripts 9.4k downloads

shapviz - SHAP Visualizations

Visualizations for SHAP (SHapley Additive exPlanations), such as waterfall plots, force plots, various types of importance plots, dependence plots, and interaction plots. These plots act on a 'shapviz' object created from a matrix of SHAP values and a corresponding feature dataset. Wrappers for the R packages 'xgboost', 'lightgbm', 'fastshap', 'shapr', 'h2o', 'treeshap', 'DALEX', and 'kernelshap' are added for convenience. By separating visualization and computation, it is possible to display factor variables in graphs, even if the SHAP values are calculated by a model that requires numerical features. The plots are inspired by those provided by the 'shap' package in Python, but there is no dependency on it.

Last updated

explainable-aimachine-learningshapshapley-valuevisualizationxai

11.15 score 120 stars 1 dependents 718 scripts 9.1k downloads

ingredients - Effects and Importances of Model Ingredients

Collection of tools for assessment of feature importance and feature effects. Key functions are: feature_importance() for assessment of global level feature importance, ceteris_paribus() for calculation of the what-if plots, partial_dependence() for partial dependence plots, conditional_dependence() for conditional dependence plots, accumulated_dependence() for accumulated local effects plots, aggregate_profiles() and cluster_profiles() for aggregation of ceteris paribus profiles, generic print() and plot() for better usability of selected explainers, generic plotD3() for interactive, D3 based explanations, and generic describe() for explanations in natural language. The package 'ingredients' is a part of the 'DrWhy.AI' universe (Biecek 2018) <arXiv:1806.08915>.

Last updated

10.67 score 38 stars 20 dependents 96 scripts 9.0k downloads

iBreakDown - Model Agnostic Instance Level Variable Attributions

Model agnostic tool for decomposition of predictions from black boxes. Supports additive attributions and attributions with interactions. The Break Down Table shows contributions of every variable to a final prediction. The Break Down Plot presents variable contributions in a concise graphical way. This package works for classification and regression models. It is an extension of the 'breakDown' package (Staniak and Biecek 2018) <doi:10.32614/RJ-2018-072>, with new and faster strategies for orderings. It supports interactions in explanations and has interactive visuals (implemented with 'D3.js' library). The methodology behind is described in the 'iBreakDown' article (Gosiewska and Biecek 2019) <arXiv:1903.11420> This package is a part of the 'DrWhy.AI' universe (Biecek 2018) <arXiv:1806.08915>.

Last updated

breakdownimlinterpretabilityshapleyxai

10.30 score 84 stars 19 dependents 68 scripts 7.6k downloads

randomForestExplainer - Explaining and Visualizing Random Forests in Terms of Variable Importance

A set of tools to help explain which variables are most important in a random forests. Various variable importance measures are calculated and visualized in different settings in order to get an idea on how their importance changes depending on our criteria (Hemant Ishwaran and Udaya B. Kogalur and Eiran Z. Gorodeski and Andy J. Minn and Michael S. Lauer (2010) <doi:10.1198/jasa.2009.tm08622>, Leo Breiman (2001) <doi:10.1023/A:1010933404324>).

Last updated

random-forest

9.44 score 240 stars 326 scripts 800 downloads

kernelshap - Kernel SHAP

Efficient implementation of Kernel SHAP (Lundberg and Lee, 2017, <doi:10.48550/arXiv.1705.07874>) permutation SHAP, and additive SHAP for model interpretability. For Kernel SHAP and permutation SHAP, if the number of features is too large for exact calculations, the algorithms iterate until the SHAP values are sufficiently precise in terms of their standard errors. The package integrates smoothly with meta-learning packages such as 'tidymodels', 'caret' or 'mlr3'. It supports multi-output models, case weights, and parallel computations. Visualizations can be done using the R package 'shapviz'.

Last updated

explainable-aiinterpretabilityinterpretable-machine-learningmachine-learningshapxai

9.43 score 62 stars 23 dependents 227 scripts 9.3k downloads

auditor - Model Audit - Verification, Validation, and Error Analysis

Provides an easy to use unified interface for creating validation plots for any model. The 'auditor' helps to avoid repetitive work consisting of writing code needed to create residual plots. This visualizations allow to asses and compare the goodness of fit, performance, and similarity of models.

Last updated

classificationerror-analysisexplainable-artificial-intelligencemachine-learningmodel-validationregression-modelsresidualsxai

9.36 score 59 stars 2 dependents 100 scripts 3.7k downloads

survex - Explainable Machine Learning in Survival Analysis

Survival analysis models are commonly used in medicine and other areas. Many of them are too complex to be interpreted by human. Exploration and explanation is needed, but standard methods do not give a broad enough picture. 'survex' provides easy-to-apply methods for explaining survival models, both complex black-boxes and simpler statistical models. They include methods specific to survival analysis such as SurvSHAP(t) introduced in Krzyzinski et al., (2023) <doi:10.1016/j.knosys.2022.110234>, SurvLIME described in Kovalev et al., (2020) <doi:10.1016/j.knosys.2020.106164> as well as extensions of existing ones described in Biecek et al., (2021) <doi:10.1201/9780429027192>.

Last updated

biostatisticsbrier-scorescensored-datacox-modelcox-regressionexplainable-aiexplainable-machine-learningexplainable-mlexplanatory-model-analysisinterpretable-machine-learninginterpretable-mlmachine-learningprobabilistic-machine-learningshapsurvival-analysistime-to-eventvariable-importancexai

8.60 score 118 stars 1 dependents 113 scripts 440 downloads

DALEXtra - Extension for 'DALEX' Package

Provides wrapper of various machine learning models. In applied machine learning, there is a strong belief that we need to strike a balance between interpretability and accuracy. However, in field of the interpretable machine learning, there are more and more new ideas for explaining black-box models, that are implemented in 'R'. 'DALEXtra' creates 'DALEX' Biecek (2018) <doi:10.48550/arXiv.1806.08915> explainer for many type of models including those created using 'python' 'scikit-learn' and 'keras' libraries, and 'java' 'h2o' library. Important part of the package is Champion-Challenger analysis and innovative approach to model performance across subsets of test data presented in Funnel Plot.

Last updated

extension-for-dalex-package

8.56 score 67 stars 1 dependents 543 scripts 5.5k downloads

modelStudio - Interactive Studio for Explanatory Model Analysis

Automate the explanatory analysis of machine learning predictive models. Generate advanced interactive model explanations in the form of a serverless HTML site with only one line of code. This tool is model-agnostic, therefore compatible with most of the black-box predictive models and frameworks. The main function computes various (instance and model-level) explanations and produces a customisable dashboard, which consists of multiple panels for plots with their short descriptions. It is possible to easily save the dashboard and share it with others. 'modelStudio' facilitates the process of Interactive Explanatory Model Analysis introduced in Baniecki et al. (2023) <doi:10.1007/s10618-023-00924-w>.

Last updated

aiexplainableexplainable-aiexplainable-machine-learningexplanatory-model-analysishumanimlinteractiveinteractivityinterpretabilityinterpretableinterpretable-machine-learninglearningmachinemodelmodel-visualizationvisualizationxai

7.97 score 334 stars 62 scripts 362 downloads

fairmodels - Flexible Tool for Bias Detection, Visualization, and Mitigation

Measure fairness metrics in one place for many models. Check how big is model's bias towards different races, sex, nationalities etc. Use measures such as Statistical Parity, Equal odds to detect the discrimination against unprivileged groups. Visualize the bias using heatmap, radar plot, biplot, bar chart (and more!). There are various pre-processing and post-processing bias mitigation algorithms implemented. Package also supports calculating fairness metrics for regression models. Find more details in (Wiśniewski, Biecek (2021)) <arXiv:2104.00507>.

Last updated

explain-classifiersexplainable-mlfairnessfairness-comparisonfairness-mlmodel-evaluation

7.87 score 85 stars 1 dependents 73 scripts 497 downloads

shapper - Wrapper of Python Library 'shap'

Provides SHAP explanations of machine learning models. In applied machine learning, there is a strong belief that we need to strike a balance between interpretability and accuracy. However, in field of the Interpretable Machine Learning, there are more and more new ideas for explaining black-box models. One of the best known method for local explanations is SHapley Additive exPlanations (SHAP) introduced by Lundberg, S., et al., (2016) <arXiv:1705.07874> The SHAP method is used to calculate influences of variables on the particular observation. This method is based on Shapley values, a technique used in game theory. The R package 'shapper' is a port of the Python library 'shap'.

Last updated

7.38 score 58 stars 69 scripts 675 downloads

treeshap - Compute SHAP Values for Your Tree-Based Models Using the 'TreeSHAP' Algorithm

An efficient implementation of the 'TreeSHAP' algorithm introduced by Lundberg et al., (2020) <doi:10.1038/s42256-019-0138-9>. It is capable of calculating SHAP (SHapley Additive exPlanations) values for tree-based models in polynomial time. Currently supported models include 'gbm', 'randomForest', 'ranger', 'xgboost', 'lightgbm'.

Last updated

explainabilityexplainable-aiexplainable-artificial-intelligenceexplanatory-model-analysisimlinterpretabilityinterpretable-machine-learningmachine-learningresponsible-mlshapshapley-valuexaicpp

7.30 score 98 stars 195 scripts 1.2k downloads

localModel - LIME-Based Explanations with Interpretable Inputs Based on Ceteris Paribus Profiles

Local explanations of machine learning models describe, how features contributed to a single prediction. This package implements an explanation method based on LIME (Local Interpretable Model-agnostic Explanations, see Tulio Ribeiro, Singh, Guestrin (2016) <doi:10.1145/2939672.2939778>) in which interpretable inputs are created based on local rather than global behaviour of each original feature.

Last updated

6.80 score 14 stars 28 scripts 3.6k downloads

rSAFE - Surrogate-Assisted Feature Extraction

Provides a model agnostic tool for white-box model trained on features extracted from a black-box model. For more information see: Gosiewska et al. (2020) <doi:10.1016/j.dss.2021.113556>.

Last updated

feature-engineeringfeature-extractionimlinterpretabilitymachine-learningxai

6.80 score 28 stars 45 scripts 233 downloads

hstats - Interaction Statistics

Fast, model-agnostic implementation of different H-statistics introduced by Jerome H. Friedman and Bogdan E. Popescu (2008) <doi:10.1214/07-AOAS148>. These statistics quantify interaction strength per feature, feature pair, and feature triple. The package supports multi-output predictions and can account for case weights. In addition, several variants of the original statistics are provided. The shape of the interactions can be explored through partial dependence plots or individual conditional expectation plots. 'DALEX' explainers, meta learners ('mlr3', 'tidymodels', 'caret') and most other models work out-of-the-box.

Last updated

interactioninterpretabilitymachine-learningrstatstatisticsxai

6.04 score 34 stars 1 dependents 71 scripts 318 downloads

arenar - Arena for the Exploration and Comparison of any ML Models

Generates data for challenging machine learning models in 'Arena' <https://arena.drwhy.ai> - an interactive web application. You can start the server with XAI (Explainable Artificial Intelligence) plots to be generated on-demand or precalculate and auto-upload data file beside shareable 'Arena' URL.

Last updated

axplainable-artificial-intelligenceemaexplainabilityexplanatory-model-analysisimlinteractive-xaiinterpretabilityxai

5.94 score 31 stars 14 scripts 255 downloads

modelDown - Make Static HTML Website for Predictive Models

Website generator with HTML summaries for predictive models. This package uses 'DALEX' explainers to describe global model behavior. We can see how well models behave (tabs: Model Performance, Auditor), how much each variable contributes to predictions (tabs: Variable Response) and which variables are the most important for a given model (tabs: Variable Importance). We can also compare Concept Drift for pairs of models (tabs: Drifter). Additionally, data available on the website can be easily recreated in current R session. Work on this package was financially supported by the NCN Opus grant 2017/27/B/ST6/01307 at Warsaw University of Technology, Faculty of Mathematics and Information Science.

Last updated

5.94 score 120 stars 18 scripts 238 downloads

EIX - Explain Interactions in 'XGBoost'

Structure mining from 'XGBoost' and 'LightGBM' models. Key functionalities of this package cover: visualisation of tree-based ensembles models, identification of interactions, measuring of variable importance, measuring of interaction importance, explanation of single prediction with break down plots (based on 'xgboostExplainer' and 'iBreakDown' packages). To download the 'LightGBM' use the following link: <https://github.com/Microsoft/LightGBM>. 'EIX' is a part of the 'DrWhy.AI' universe.

Last updated

5.72 score 26 stars 5 scripts 604 downloads

live - Local Interpretable (Model-Agnostic) Visual Explanations

Interpretability of complex machine learning models is a growing concern. This package helps to understand key factors that drive the decision made by complicated predictive model (so called black box model). This is achieved through local approximations that are either based on additive regression like model or CART like model that allows for higher interactions. The methodology is based on Tulio Ribeiro, Singh, Guestrin (2016) <doi:10.1145/2939672.2939778>. More details can be found in Staniak, Biecek (2018) <doi:10.32614/RJ-2018-072>.

Last updated

imlinterpretabilitylimemachine-learningmodel-visualizationvisual-explanationsxai

5.59 score 35 stars 55 scripts 234 downloads

vivo - Variable Importance via Oscillations

Provides an easy to calculate local variable importance measure based on Ceteris Paribus profile and global variable importance measure based on Partial Dependence Profiles.

Last updated

explainable-aiexplainable-artificial-intelligenceexplainable-mlimlinterpretable-machine-learningvariable-importancexai

5.45 score 14 stars 7 scripts 252 downloads

drifter - Concept Drift and Concept Shift Detection for Predictive Models

Concept drift refers to the change in the data distribution or in the relationships between variables over time. 'drifter' calculates distances between variable distributions or variable relations and identifies both types of drift. Key functions are: calculate_covariate_drift() checks distance between corresponding variables in two datasets, calculate_residuals_drift() checks distance between residual distributions for two models, calculate_model_drift() checks distance between partial dependency profiles for two models, check_drift() executes all checks against drift. 'drifter' is a part of the 'DrWhy.AI' universe (Biecek 2018) <arXiv:1806.08915>.

Last updated

concept-driftconcept-shiftdrwhypredictive-modeling

4.45 score 19 stars 1 dependents 5 scripts 244 downloads