Falling Rule Lists

By Fulton Wang, Cynthia Rudin. 2014

265 citations recorded by the source corpus.

Falling rule lists are classification models consisting of an ordered list of if-then rules, where (i) the order of rules determines which example should be classified by each rule, and (ii) the estimated probability of success decreases monotonically down the list. These kinds of rule lists are inspired by healthcare applications where patients would be stratified into risk sets and the highest at-risk patients should be considered first. We provide a Bayesian framework for learning falling rule lists that does not rely on traditional greedy decision tree learning methods.

Subfields: Statistical Learning Theory

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistical Learning Theory

Citations in the corpus

Interpretable Machine Learning: Fundamental Principles and 10 Grand Challenges (2021, background): “Sometimes decision lists or decision sets are optimized by sampling [180, 321, 306], providing a Bayesian interpretation.”
Stop Explaining Black Box Machine Learning Models for High Stakes Decisions and Use Interpretable Models Instead (2018, methodology): “s to create algorithms that solve logical modeling problems in a reasonable amount of time, for practical datasets. We have been extending CORELS to more complex problems, such as Falling Rule Lists [Wang and Rudin, 2015, Chen and Rudin, 2018], and optimal binary-split decision trees, but there is m”
Please Stop Explaining Black Box Models for High Stakes Decisions (2018, methodology): “We have been extending CORELS to more complex problems, such as Falling Rule Lists [Wang and Rudin, 2015, Chen and Rudin, 2018], and optimal binary-split decision trees, but there is much work to be done on other types of logical models, with various kinds of constraints.”
Local Rule-Based Explanations of Black Box Decision Systems (2018, background): “ing set is available, classification rules are also widely used to avoid building black boxes by directly designing a transparent classifier [17] which is locally or globally interpretable on its own [23, 25, 39]. Other approaches, more related to the one we propose, address the problem of explainin”
Anchors: High-Precision Model-Agnostic Explanations (2018, background): “As a consequence of the need for users to understand the behavior of these models, interpretable machine learning has seen a resurgence in recent years, ranging from the design of novel globally-interpretable machine learning models (Lakkaraju, Bach, and Leskovec 2016; Ustun and Rudin 2015; Wang and”
A Survey of Methods for Explaining Black Box Models (2018, methodology): “3], [51], [97] Outcome Explanation [108], [25], [113], [89], [56], [83], [98] Black Box Inspection [71], [7], [92], [16], [35], [30], [48], [2], [1], [110], [96], [90], [81] Transparent Design [109], [102], [57], [91], [53], [104], [64], [8], [44], [63], [47], [103] Open The Black Box Survey 39 A Su”
Interpretability via Model Extraction (2017, background): “Complimentary approaches to interpretability focus on learning interpretable models [7, 26, 30] or on explaining the model’s behavior on specific inputs rather than the model as a whole [23].”
Model-Agnostic Interpretability of Machine Learning (2016, methodology): “The prevailing solution to this explanation problem is to use so called “interpretable” models, such as decision trees, rules (Letham et al., 2015; Wang & Rudin, 2015), additive models (Caruana et al., 2015), attention-based networks (Xu et al., 2015), or sparse linear models (Ustun & Rudin, 2015).”
The Promise and Peril of Human Evaluation for Model Interpretability (2017, background): “rify different components of interpretable explanation model systems. An explanation vehicle is themodelclass ofan explanationsuchas decisiontrees, generalized additive models [3], falling rule lists [17]. We expand this deﬁnition to pairings of model class and a ﬁxed explanation complexity constrai”
Learning Certifiably Optimal Rule Lists (2017, background): “For example, Falling Rule Lists (Wang and Rudin, 2015) are constrained to have decreasing probabilities down the list as are rule lists for dynamic treat-ment regimes (Zhang et al., 2015) and cost-sensitive dynamic treatment regimes (Lakkaraju and Rudin, 2017).”
Interpretable classification models for recidivism prediction (2015, methodology): “In Appendix C, we present accurate and interpretable models that were created using a new rule-based method known as Falling Rule Lists.”
Evaluating Reinforcement Learning Algorithms in Observational Health Settings (2018, neutral): “…addressed the issue of modifying ML algorithms to produce more interpretable predictions [Vellido et al., 2012, Doshi-Velez and Kim, 2017, Lipton, 2016, Wang and Rudin, 2015] and some have tackled the speciﬁc problem of interpretability of ML in the healthcare domain [Caruana et al., 2015].”
Learning Fair Rule Lists (2019, background): “For instance, when they have small or of reasonable size [44], ∗ Principal contributor. the following models can be considered as interpretable: rules sets [18, 43, 46, 54], rule lists [3, 59, 61], decision trees [9, 51] and scoring systems [39, 57, 64].”
Deep Weighted Averaging Classifiers (2018, background): “This category includes classic models that can easily be simulated by humans, such as decision lists, as well as sparse linear models, where the prediction is based on a weighted sum of features [6, 24, 29, 49, 50, 56].”
Learning customized and optimized lists of rules with mathematical programming (2018, methodology): “ABayesian counterpart to thework of Goh andRudin [26] is that ofWang et al [71], which proves some results analogous to those in Sect.”
Boolean Decision Rules via Column Generation (2018, background): “Rule sets stand in contrast to decision lists [44, 35, 49, 3, 34, 53], where rules are ordered in an IF-ELSE sequence, and decision trees [11, 43, 6], where they are organized into a tree structure.”
Manipulating and Measuring Model Interpretability (2018, neutral): “Malignancy risk in mammography Age tumor density, 5 binary variables describing tumor shape Full mammogram image, full medical records, clinical interview (habits, family history, etc), plus 10 binary variables not in interpretable model [111]”