Interpretable Rule Discovery Through Bilevel Optimization of Split-Rules of Nonlinear Decision Trees for Classification Problems

Dhebar, Yashesh; Deb, Kalyanmoy

doi:10.1109/tcyb.2020.3033003

Cited by 16 publications

(9 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Rule extraction for interpretability purposes is very common in tree-based approaches [36][37][38] , however rule extraction from neural networks is rare, with the majority of methods being limited to rule extraction from networks with one hidden layer 18 . Zilke et al 39 extract rules from networks with more than one hidden layer (i.e., DNNs), however, the approach has scalability issues due to prohibitively high memory and time consumption 4,39 .…”

Section: Discussionmentioning

confidence: 99%

REM: An Integrative Rule Extraction Methodology for Explainable Data Analysis in Healthcare

Shams

Dimanov

et al. 2021

Preprint

View full text Add to dashboard Cite

Deep learning models are receiving increasing attention in clinical decision-making, however the lack of interpretability and explainability impedes their deployment in day-to-day clinical practice. We propose REM, an interpretable and explainable methodology for extracting rules from deep neural networks and combining them with other data-driven and knowledge-driven rules. This allows integrating machine learning and reasoning for investigating applied and basic biological research questions. We evaluate the utility of REM on the predictive tasks of classifying histological and immunohistochemical breast cancer subtypes from genotype and phenotype data. We demonstrate that REM efficiently extracts accurate, comprehensible and, biologically relevant rulesets from deep neural networks that can be readily integrated with rulesets obtained from tree-based approaches. REM provides explanation facilities for predictions and enables the clinicians to validate and calibrate the extracted rulesets with their domain knowledge. With these functionalities, REM caters for a novel and direct human-in-the-loop approach in clinical decision making.

show abstract

Section: Discussionmentioning

confidence: 99%

REM: An Integrative Rule Extraction Methodology for Explainable Data Analysis in Healthcare

Shams

Dimanov

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Recently, an evolutionary algorithm based non-linear decision tree classifier was proposed in [8]. The classifier is represented in the form of a non-linear decision tree as shown in Figure 7.…”

Section: E Nonlinear Decision Tree (Nldt) Approachmentioning

confidence: 99%

“…• Training is slower as compared to CART and SVM. Details regarding the bilevel optimization algorithm and parameter settings can be found from [8].…”

Section: Advantagesmentioning

confidence: 99%

“…A. Customized Data: DS1-4 and modified DS1-3 Four synthetic two dimensional datasets DS1-DS4 and their variants m-DS1, m-DS2 and m-DS3 are generated using the procedure provied in [8] to investigate behavior of classification algorithms across following properties:…”

Section: Datasets Consideredmentioning

confidence: 99%

“…1) Test problems: We use modified versions of ZDT [23] and DTLZ [7] problems with two and three objectives, respectively to generate datasets involving 30 and 500 features (details in [8]). These two problem sizes also allow us to perform a scale-up study of the classification methods.…”

Section: Pareto Versus Non-pareto Classificationmentioning

confidence: 99%

See 2 more Smart Citations

Evaluating Nonlinear Decision Trees for Binary Classification Tasks with Other Existing Methods

Dhebar

Gupta

Deb

2020

2020 IEEE Symposium Series on Computational Intelligence (SSCI)

Self Cite

View full text Add to dashboard Cite

Classification of datasets into two or more distinct classes is an important machine learning task. Many methods are able to classify binary classification tasks with a very high accuracy on test data, but cannot provide any easily interpretable explanation for users to have a deeper understanding of reasons for the split of data into two classes. In this paper, we highlight and evaluate a recently proposed nonlinear decision tree approach with a number of commonly used classification methods on a number of datasets involving a few to a large number of features. The study reveals key issues such as effect of classification on the method's parameter values, complexity of the classifier versus achieved accuracy, and interpretability of resulting classifiers.

show abstract

Geometrically interpretable Variance Hyper Rectangle learning for pattern classification

Sun

Peng

et al. 2022

Engineering Applications of Artificial Intelligence

View full text Add to dashboard Cite

Interpretable Rule Discovery Through Bilevel Optimization of Split-Rules of Nonlinear Decision Trees for Classification Problems

Cited by 16 publications

References 29 publications

REM: An Integrative Rule Extraction Methodology for Explainable Data Analysis in Healthcare

REM: An Integrative Rule Extraction Methodology for Explainable Data Analysis in Healthcare

Evaluating Nonlinear Decision Trees for Binary Classification Tasks with Other Existing Methods

Geometrically interpretable Variance Hyper Rectangle learning for pattern classification

Contact Info

Product

Resources

About