Python package for concise, transparent, and accurate predictive modeling.
All sklearn-compatible and easy to use.
Check out our new packages! Interpretability in text: imodelsX, interpretability tools for tabular data with agents: agentic-imodels

πŸ“š docs β€’ πŸ“– demo notebooks

Modern machine-learning models are increasingly complex, often making them difficult to interpret. This package provides a simple interface for fitting and using state-of-the-art interpretable models, all compatible with scikit-learn. These models can often replace black-box models (e.g. random forests) with simpler models (e.g. rule lists) while improving interpretability and computational efficiency, all without sacrificing predictive accuracy! Simply import a classifier or regressor and use the `fit` and `predict` methods, same as standard scikit-learn models. ```python from imodels import get_clean_dataset, HSTreeClassifierCV # import any imodels model here from sklearn.model_selection import train_test_split # prepare data (a sample clinical dataset) X, y, feature_names = get_clean_dataset('csi_pecarn_pred') X_train, X_test, y_train, y_test = train_test_split( X, y, random_state=42) # fit the model model = HSTreeClassifierCV(max_leaf_nodes=4) # initialize a tree model and specify only 4 leaf nodes model.fit(X_train, y_train, feature_names=feature_names) # fit model preds = model.predict(X_test) # discrete predictions: shape is (n_test, 1) preds_proba = model.predict_proba(X_test) # predicted probabilities: shape is (n_test, n_classes) print(model) # print the model ``` ``` ------------------------------ Decision Tree with Hierarchical Shrinkage Prediction is made by looking at the value in the appropriate leaf of the tree ------------------------------ |--- FocalNeuroFindings2 <= 0.50 | |--- HighriskDiving <= 0.50 | | |--- Torticollis2 <= 0.50 | | | |--- value: [0.10] | | |--- Torticollis2 > 0.50 | | | |--- value: [0.30] | |--- HighriskDiving > 0.50 | | |--- value: [0.68] |--- FocalNeuroFindings2 > 0.50 | |--- value: [0.42] ``` ### Installation Install with `pip install imodels` (see [here](https://github.com/csinva/imodels/blob/master/docs/troubleshooting.md) for help). ### Supported models

πŸ—‚οΈ Docs   πŸ“„ Research paper   πŸ”— Reference code implementation

| Model | Reference | Description | | :-------------------------- | ------------------------------------------------------------ | ------------------------------------------------------------ | | Rulefit rule set | [πŸ—‚οΈ](https://csinva.io/imodels/rule_set/rule_fit.html), [πŸ“„](http://statweb.stanford.edu/~jhf/ftp/RuleFit.pdf), [πŸ”—](https://github.com/christophM/rulefit) | Fits a sparse linear model on rules extracted from decision trees | | Skope rule set | [πŸ—‚οΈ](https://csinva.io/imodels/rule_set/skope_rules.html#imodels.rule_set.skope_rules.SkopeRulesClassifier), [πŸ”—](https://github.com/scikit-learn-contrib/skope-rules) | Extracts rules from gradient-boosted trees, deduplicates them,
then linearly combines them based on their OOB precision | | Boosted rule set | [πŸ—‚οΈ](https://csinva.io/imodels/rule_set/boosted_rules.html), [πŸ“„](https://www.sciencedirect.com/science/article/pii/S002200009791504X), [πŸ”—](https://github.com/jaimeps/adaboost-implementation) | Sequentially fits a set of rules with Adaboost | | Slipper rule set | [πŸ—‚οΈ](https://csinva.io/imodels/rule_set/slipper.html), [πŸ“„](https://www.aaai.org/Papers/AAAI/1999/AAAI99-049.pdf) | Sequentially learns a set of rules with SLIPPER | | Bayesian rule set | [πŸ—‚οΈ](https://csinva.io/imodels/rule_set/brs.html#imodels.rule_set.brs.BayesianRuleSetClassifier), [πŸ“„](https://www.jmlr.org/papers/volume18/16-003/16-003.pdf), [πŸ”—](https://github.com/wangtongada/BOA) | Finds concise rule set with Bayesian sampling (slow) | | Bayesian rule list | [πŸ—‚οΈ](https://csinva.io/imodels/rule_list/bayesian_rule_list/bayesian_rule_list.html#imodels.rule_list.bayesian_rule_list.bayesian_rule_list.BayesianRuleListClassifier), [πŸ“„](https://projecteuclid.org/journals/annals-of-applied-statistics/volume-9/issue-3/Interpretable-classifiers-using-rules-and-Bayesian-analysis--Building-a/10.1214/15-AOAS848.full), [πŸ”—](https://github.com/tmadl/sklearn-expertsys) | Fits compact rule list distribution with Bayesian sampling (slow) | | Greedy rule list | [πŸ—‚οΈ](https://csinva.io/imodels/rule_list/greedy_rule_list.html), [πŸ”—](https://medium.com/@penggongting/implementing-decision-tree-from-scratch-in-python-c732e7c69aea) | Uses CART to fit a list (only a single path), rather than a tree | | Fast-and-frugal tree | [πŸ—‚οΈ](https://csinva.io/imodels/rule_list/fast_frugal_tree.html) [πŸ”—](https://github.com/fasttrees/fasttrees)| | One cue per level, each able to decide | | OneR rule list | [πŸ—‚οΈ](https://csinva.io/imodels/rule_list/one_r.html), [πŸ“„](https://link.springer.com/article/10.1023/A:1022631118932) | Fits rule list restricted to only one feature | | Greedy rule tree | [πŸ—‚οΈ](https://csinva.io/imodels/tree/cart_wrapper.html), [πŸ“„](https://www.taylorfrancis.com/books/mono/10.1201/9781315139470/classification-regression-trees-leo-breiman-jerome-friedman-richard-olshen-charles-stone), [πŸ”—](https://scikit-learn.org/stable/modules/tree.html) | Greedily fits tree using CART | | C4.5 rule tree | [πŸ—‚οΈ](https://csinva.io/imodels/tree/c45_tree/c45_tree.html#imodels.tree.c45_tree.c45_tree.C45TreeClassifier), [πŸ“„](https://link.springer.com/article/10.1007/BF00993309), [πŸ”—](https://github.com/RaczeQ/scikit-learn-C4.5-tree-classifier) | Greedily fits tree using C4.5 | | TAO rule tree | [πŸ—‚οΈ](https://csinva.io/imodels/tree/tao.html), [πŸ“„](https://proceedings.neurips.cc/paper/2018/hash/185c29dc24325934ee377cfda20e414c-Abstract.html) | Fits tree using alternating optimization | | Sparse integer
linear model | [πŸ—‚οΈ](https://csinva.io/imodels/algebraic/slim.html), [πŸ“„](https://link.springer.com/article/10.1007/s10994-015-5528-6) | Sparse linear model with integer coefficients | | Tree GAM | [πŸ—‚οΈ](https://csinva.io/imodels/algebraic/tree_gam.html), [πŸ“„](https://dl.acm.org/doi/abs/10.1145/2339530.2339556), [πŸ”—](https://github.com/interpretml/interpret) | Generalized additive model fit with short boosted trees | | Greedy tree
sums (FIGS) | [πŸ—‚οΈ](https://csinva.io/imodels/figs.html),γ…€[πŸ“„](https://arxiv.org/abs/2201.11931) | Sum of small trees with very few total rules (FIGS) | | Hierarchical
shrinkage wrapper | [πŸ—‚οΈ](https://csinva.io/imodels/shrinkage.html), [πŸ“„](https://arxiv.org/abs/2202.00858) | Improve a decision tree, random forest, or
gradient-boosting ensemble with ultra-fast, post-hoc regularization | | RF+ (MDI+) | [πŸ—‚οΈ](https://csinva.io/imodels/mdi_plus.html), [πŸ“„](https://arxiv.org/pdf/2307.01932) | Flexible random forest-based feature importance | | Distillation
wrapper | [πŸ—‚οΈ](https://csinva.io/imodels/util/distillation.html) | Train a black-box model,
then distill it into an interpretable model | | AutoML wrapper | [πŸ—‚οΈ](https://csinva.io/imodels/util/automl.html) | Automatically fit and select an interpretable model | | More models | βŒ› | (Coming soon!) Lightweight Rule Induction, MLRules, ... | ## Demo notebooks Demos are contained in the [notebooks](notebooks) folder
Quickstart demo Shows how to fit, predict, and visualize with different interpretable models
Autogluon demo Fit/select an interpretable model automatically using Autogluon AutoML
Clinical decision rule notebook Shows an example of using imodels for deriving a clinical decision rule
Posthoc analysis We also include some demos of posthoc analysis, which occurs after fitting models: posthoc.ipynb shows different simple analyses to interpret a trained model and uncertainty.ipynb contains basic code to get uncertainty estimates for a model
## What's the difference between the models? The final form of the above models takes one of the following forms, which aim to be simultaneously simple to understand and highly predictive: | Rule set | Rule list | Rule tree | Algebraic models | | :----------------------------------------------------------: | :-----------------------------------------------------: | :-----------------------------------------------------: | :----------------------------------------------------------: | | | | | | Different models and algorithms vary not only in their final form but also in different choices made during modeling, such as how they generate, select, and postprocess rules: | Rule candidate generation | Rule selection | Rule postprocessing| | :----------------------------------------------------------: | :--------------------------------------------------------: | :-------------------------------------------------------: | | | | |
Ex. RuleFit vs. SkopeRules RuleFit and SkopeRules differ only in the way they prune rules: RuleFit uses a linear model whereas SkopeRules heuristically deduplicates rules sharing overlap.
Ex. Bayesian rule lists vs. greedy rule lists Bayesian rule lists and greedy rule lists differ in how they select rules; bayesian rule lists perform a global optimization over possible rule lists while Greedy rule lists pick splits sequentially to maximize a given criterion.
Ex. FPSkope vs. SkopeRules FPSkope and SkopeRules differ only in the way they generate candidate rules: FPSkope uses FPgrowth whereas SkopeRules extracts rules from decision trees.

Support for different tasks

 Different models support different machine-learning tasks. Current support for different models is given below (each of these models can be imported directly from imodels (e.g. `from imodels import RuleFitClassifier`): All of these models follow the standard sklearn estimator API, which is checked for every model in [tests/model_api_test.py](tests/model_api_test.py): `fit` returns the estimator, `predict` returns labels drawn from `classes_` (strings included), `predict_proba` returns an `(n_samples, n_classes)` matrix whose rows sum to 1, DataFrame input sets `feature_names_in_`,, models can be `clone`d and configured with `get_params`/`set_params`, and every model works inside sklearn pipelines and grid searches. | Model | Binary classification | Regression | Notes | | :-------------------------- | :----------------------------------------------------------: | :----------------------------------------------------------: | --------------------------- | | Rulefit rule set | [RuleFitClassifier](https://csinva.io/imodels/rule_set/rule_fit.html#imodels.rule_set.rule_fit.RuleFitClassifier) | [RuleFitRegressor](https://csinva.io/imodels/rule_set/rule_fit.html#imodels.rule_set.rule_fit.RuleFitRegressor) | | | Skope rule set | [SkopeRulesClassifier](https://csinva.io/imodels/rule_set/skope_rules.html#imodels.rule_set.skope_rules.SkopeRulesClassifier) | | | | FPSkope rule set | [FPSkopeClassifier](https://csinva.io/imodels/rule_set/fpskope.html#imodels.rule_set.fpskope.FPSkopeClassifier) | | Like Skope, but generates candidate rules with FPGrowth; requires discretized features | | Rulefit rule set (XGBoost) | pass `tree_generator=XGBClassifier(...)` to `RuleFitClassifier` | pass `tree_generator=XGBRegressor(...)` to `RuleFitRegressor` | Requires [xgboost](https://pypi.org/project/xgboost/) | | FPLasso rule set | [FPLassoClassifier](https://csinva.io/imodels/rule_set/fplasso.html#imodels.rule_set.fplasso.FPLassoClassifier) | [FPLassoRegressor](https://csinva.io/imodels/rule_set/fplasso.html#imodels.rule_set.fplasso.FPLassoRegressor) | Lasso over rules mined with FPGrowth; requires discretized features | | Boosted rule set | [BoostedRulesClassifier](https://csinva.io/imodels/rule_set/boosted_rules.html#imodels.rule_set.boosted_rules.BoostedRulesClassifier) | [BoostedRulesRegressor](https://csinva.io/imodels/rule_set/boosted_rules.html#imodels.rule_set.boosted_rules.BoostedRulesRegressor) | | | SLIPPER rule set | [SlipperClassifier](https://csinva.io/imodels/rule_set/slipper.html#imodels.rule_set.slipper.SlipperClassifier) | | | | Bayesian rule set | [BayesianRuleSetClassifier](https://csinva.io/imodels/rule_set/brs.html#imodels.rule_set.brs.BayesianRuleSetClassifier) | | Fails for large problems | | Bayesian rule list | [BayesianRuleListClassifier](https://csinva.io/imodels/rule_list/bayesian_rule_list/bayesian_rule_list.html#imodels.rule_list.bayesian_rule_list.bayesian_rule_list.BayesianRuleListClassifier) | | | | Greedy rule list | [GreedyRuleListClassifier](https://csinva.io/imodels/rule_list/greedy_rule_list.html#imodels.rule_list.greedy_rule_list.GreedyRuleListClassifier) | | | | OneR rule list | [OneRClassifier](https://csinva.io/imodels/rule_list/one_r.html#imodels.rule_list.one_r.OneRClassifier) | | | | Greedy rule tree (CART) | [GreedyTreeClassifier](https://csinva.io/imodels/tree/cart_wrapper.html#imodels.tree.cart_wrapper.GreedyTreeClassifier) | [GreedyTreeRegressor](https://csinva.io/imodels/tree/cart_wrapper.html#imodels.tree.cart_wrapper.GreedyTreeRegressor) | | | C4.5 rule tree | [C45TreeClassifier](https://csinva.io/imodels/tree/c45_tree/c45_tree.html#imodels.tree.c45_tree.c45_tree.C45TreeClassifier) | | | | CCP-pruned rule tree | [DecisionTreeCCPClassifier](https://csinva.io/imodels/tree/cart_ccp.html#imodels.tree.cart_ccp.DecisionTreeCCPClassifier) | [DecisionTreeCCPRegressor](https://csinva.io/imodels/tree/cart_ccp.html#imodels.tree.cart_ccp.DecisionTreeCCPRegressor) | Prunes a tree to a target complexity via cost-complexity pruning | | TAO rule tree | [TaoTreeClassifier](https://csinva.io/imodels/tree/tao.html#imodels.tree.tao.TaoTreeClassifier) | [TaoTreeRegressor](https://csinva.io/imodels/tree/tao.html#imodels.tree.tao.TaoTreeRegressor) | | | Sparse integer linear model | [SLIMClassifier](https://csinva.io/imodels/algebraic/slim.html#imodels.algebraic.slim.SLIMClassifier) | [SLIMRegressor](https://csinva.io/imodels/algebraic/slim.html#imodels.algebraic.slim.SLIMRegressor) | Requires extra dependencies for speed | | Tree GAM | [TreeGAMClassifier](https://csinva.io/imodels/algebraic/tree_gam.html) | [TreeGAMRegressor](https://csinva.io/imodels/algebraic/tree_gam.html) | | | Greedy tree sums (FIGS) | [FIGSClassifier](https://csinva.io/imodels/tree/figs.html#imodels.tree.figs.FIGSClassifier) | [FIGSRegressor](https://csinva.io/imodels/tree/figs.html#imodels.tree.figs.FIGSRegressor) | | | Hierarchical shrinkage | [HSTreeClassifierCV](https://csinva.io/imodels/tree/hierarchical_shrinkage.html#imodels.tree.hierarchical_shrinkage.HSTreeClassifierCV) | [HSTreeRegressorCV](https://csinva.io/imodels/tree/hierarchical_shrinkage.html#imodels.tree.hierarchical_shrinkage.HSTreeRegressorCV) | Wraps any sklearn tree-based model | | Marginal shrinkage
linear model | | [MarginalShrinkageLinearModelRegressor](https://csinva.io/imodels/algebraic/marginal_shrinkage_linear_model.html) | Linear model shrunk towards its marginal effects | | BART | | [BART](https://csinva.io/imodels/experimental/bartpy/index.html) | Bayesian additive regression trees (slow) | | Distillation | | [DistilledRegressor](https://csinva.io/imodels/util/distillation.html#imodels.util.distillation.DistilledRegressor) | Wraps any sklearn-compatible models | | AutoML model | [AutoInterpretableClassifier️](https://csinva.io/imodels/util/automl.html) | [AutoInterpretableRegressor️](https://csinva.io/imodels/util/automl.html) | | **Feature scaling.** Most models here work on raw features. `SLIMClassifier` and `SLIMRegressor` are the exception: their coefficients are integers, so features on very different scales collapse to zero when rounded. Standardize X before fitting them (they warn if rounding has removed most of the model). **Multiclass.** These classifiers handle more than two classes: `FIGSClassifier`, `GreedyTreeClassifier`, `HSTreeClassifier`, `TaoTreeClassifier`, `BoostedRulesClassifier`, `SLIMClassifier`, `C45TreeClassifier`, `DecisionTreeCCPClassifier` and the `CV` variants. The rule-set and rule-list models are binary-only and raise a clear error if given a multiclass target, rather than silently treating it as binary. **Categorical features.** `FIGS` takes them directly β€” pass the column names and it one-hot encodes them internally, remembering them for `predict`: ```python model = FIGSClassifier().fit(X, y, categorical_features=['pet', 'city']) model.predict(X) ``` Other models expect numeric input, so encode categorical columns first (e.g. with `sklearn.preprocessing.OneHotEncoder`, or one of the [discretizers](https://csinva.io/imodels/discretization/index.html) for numeric columns that a rule model needs binarized).

Plotting trees with dtreeviz

 Tree-based models can be drawn with [dtreeviz](https://github.com/parrt/dtreeviz). `shadow_tree` builds the `ShadowDecTree` it needs from any imodels tree model: ```python import dtreeviz from imodels import FIGSClassifier, shadow_tree model = FIGSClassifier(max_rules=6).fit(X, y) viz = dtreeviz.trees.DTreeVizAPI(shadow_tree(model, X, y)) viz.view() ``` For a model made of several trees (FIGS, boosted rules), pass `tree_num` to pick one. Feature and class names default to those the model was fitted with. dtreeviz is not a dependency and is imported only when this is called.

Inspecting the rules a model learned

 Every rule-based model exposes its rules the same way, as a `pandas` DataFrame with one row per rule, via `get_rules()`: ```python from imodels import FIGSClassifier model = FIGSClassifier(max_rules=4).fit(X_train, y_train, feature_names=feature_names) model.get_rules() ``` ``` rule prediction tree 0 FocalNeuroFindings2 <= 0.5 0.117 0 1 FocalNeuroFindings2 > 0.5 0.427 0 2 HighriskDiving <= 0.5 -0.008 1 3 HighriskDiving > 0.5 0.550 1 4 PainNeck2 <= 0.5 and AlteredMentalStatus2 <= 0.5 -0.083 2 5 PainNeck2 <= 0.5 and AlteredMentalStatus2 > 0.5 0.048 2 6 PainNeck2 > 0.5 0.058 2 ``` Two columns are always present: `rule`, the condition as a string, and `prediction`, what that rule predicts. Models add their own columns on top β€” `coef`, `support` and `importance` for RuleFit, `tree` for models made of several trees, and `weight` for boosted ensembles, which combine their trees by weighted vote. Where a model is additive, as FIGS is, `prediction` is that tree's contribution, so the contributions of the matching rules sum to the model's output. This works across rule sets, rule lists and tree-based models (RuleFit, SkopeRules, SLIPPER, greedy and Bayesian rule lists, FIGS, CART, C4.5, TAO, boosted rules, and hierarchical shrinkage, including the `CV` variants). It is also available as a function, `imodels.get_rules(model)`, and takes an optional `feature_names` argument to rename the features. Models that aren't rule-based raise a clear error.

SHAP values for shrunk trees

 `shap.TreeExplainer` dispatches on the model class, so it doesn't recognize the imodels wrapper. Pass the shrunk estimator it wraps: ```python import shap from imodels import HSTreeClassifier model = HSTreeClassifier(DecisionTreeClassifier(max_leaf_nodes=8), reg_param=50).fit(X, y) explainer = shap.TreeExplainer(model.estimator_) # not model itself shap_values = explainer.shap_values(X) ``` Hierarchical shrinkage rewrites the node values of that tree in place, so the explainer sees the shrunk model: the SHAP values differ from the unshrunk tree's and sum, with the expected value, to `model.predict_proba(X)`. This reproduces the SHAP summary plots in the [paper](https://arxiv.org/abs/2202.00858). Tree-based models expose `feature_importances_` (mean decrease in impurity), the same measure sklearn's tree models report, so they can be compared directly. Tree-based models also expose `apply(X)`, which reports which leaf each sample falls into, using the same node numbering as scikit-learn. A single tree returns one index per sample; a model made of several trees (FIGS, boosted rules) returns one column per tree, like `RandomForest.apply`.
### Extras
Data-wrangling functions for working with popular tabular datasets (e.g. compas). These functions, in conjunction with imodels-data and imodels-experiments, make it simple to download data and run experiments on new models.
Explain classification errors with a simple posthoc function. Fit an interpretable model to explain a previous model's errors (ex. in this notebookπŸ““).
Fast and effective discretizers for data preprocessing.
Discretizer Reference Description
MDLP πŸ—‚οΈ, πŸ”—, πŸ“„ Discretize using entropy minimization heuristic
Simple πŸ—‚οΈ, πŸ”— Simple KBins discretization
Random Forest πŸ—‚οΈ Discretize into bins based on random forest split popularity
Rule-based utils for customizing models The code here contains many useful and customizable functions for rule-based learning in the util folder. This includes functions / classes for rule deduplication, rule screening, and converting between trees, rulesets, and neural networks.
## Our favorite models After developing and playing with `imodels`, we developed a few new models to overcome limitations of existing interpretable models. ### FIGS: Fast interpretable greedy-tree sums [πŸ“„ Paper](https://arxiv.org/abs/2201.11931), [πŸ”— Post](https://csinva.io/imodels/figs.html), [πŸ“Œ Citation](https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=fast+interpretable+greedy-tree+sums&oq=fast#d=gs_cit&u=%2Fscholar%3Fq%3Dinfo%3ADnPVL74Rop0J%3Ascholar.google.com%2F%26output%3Dcite%26scirp%3D0%26hl%3Den) Fast Interpretable Greedy-Tree Sums (FIGS) is an algorithm for fitting concise rule-based models. Specifically, FIGS generalizes CART to simultaneously grow a flexible number of trees in a summation. The total number of splits across all the trees can be restricted by a pre-specified threshold, keeping the model interpretable. Experiments across a wide array of real-world datasets show that FIGS achieves state-of-the-art prediction performance when restricted to just a few splits (e.g. less than 20).

Example FIGS model. FIGS learns a sum of trees with a flexible number of trees; to make its prediction, it sums the result from each tree.

### Hierarchical shrinkage: post-hoc regularization for tree-based methods [πŸ“„ Paper](https://arxiv.org/abs/2202.00858) (ICML 2022), [πŸ”— Post](https://csinva.io/imodels/shrinkage.html), [πŸ“Œ Citation](https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=hierarchical+shrinkage+singh&btnG=&oq=hierar#d=gs_cit&u=%2Fscholar%3Fq%3Dinfo%3Azc6gtLx-aL4J%3Ascholar.google.com%2F%26output%3Dcite%26scirp%3D0%26hl%3Den) Hierarchical shrinkage is an extremely fast post-hoc regularization method which works on any decision tree (or tree-based ensemble, such as Random Forest). It does not modify the tree structure, and instead regularizes the tree by shrinking the prediction over each node towards the sample means of its ancestors (using a single regularization parameter). Experiments over a wide variety of datasets show that hierarchical shrinkage substantially increases the predictive performance of individual decision trees and decision-tree ensembles.

HS Example. HS applies post-hoc regularization to any decision tree by shrinking each node towards its parent.

### MDI+: Flexible Tree-Based Feature Importance [πŸ“„ Paper](https://arxiv.org/pdf/2307.01932.pdf), [πŸ”— Post](https://csinva.io/imodels/mdi_plus.html), [πŸ“Œ Citation](https://scholar.google.com/scholar?hl=en&as_sdt=0%2C23&q=MDI%2B%3A+A+Flexible+Random+Forest-Based+Feature+Importance+Framework&btnG=#d=gs_cit&t=1690399844081&u=%2Fscholar%3Fq%3Dinfo%3Axc0LcHXE_lUJ%3Ascholar.google.com%2F%26output%3Dcite%26scirp%3D0%26hl%3Den) MDI+ is a novel feature importance framework, which generalizes the popular mean decrease in impurity (MDI) importance score for random forests. At its core, MDI+ expands upon a recently discovered connection between linear regression and decision trees. In doing so, MDI+ enables practitioners to (1) tailor the feature importance computation to the data/problem structure and (2) incorporate additional features or knowledge to mitigate known biases of decision trees. In both real data case studies and extensive real-data-inspired simulations, MDI+ outperforms commonly used feature importance measures (e.g., MDI, permutation-based scores, and TreeSHAP) by substantional margins. ## References
Readings
Reference implementations (also linked above) The code here heavily derives from the wonderful work of previous projects. We seek to to extract out, unify, and maintain key parts of these projects.
Related packages
Updates
Please cite the package if you use it in an academic work :) ```r @software{ singh2021imodels, title = {imodels: a python package for fitting interpretable models}, journal = {Journal of Open Source Software}, publisher = {The Open Journal}, year = {2021}, author = {Singh, Chandan and Nasseri, Keyan and Tan, Yan Shuo and Tang, Tiffany and Yu, Bin}, volume = {6}, number = {61}, pages = {3192}, doi = {10.21105/joss.03192}, url = {https://doi.org/10.21105/joss.03192}, } ```