Interpretable Model

Interpretable models aim to create machine learning systems whose decision-making processes are transparent and understandable to humans, addressing the "black box" problem of many high-performing models. Current research focuses on developing inherently interpretable architectures like generalized additive models (GAMs), decision trees, rule lists, and symbolic regression, as well as post-hoc explanation methods for existing models, such as SHAP and LIME. This emphasis on interpretability is driven by the need for trust, accountability, and the ability to gain insights from complex data in fields ranging from healthcare and finance to scientific discovery, where understanding model decisions is crucial for effective application and responsible use. The development of more accurate and efficient methods for creating and evaluating interpretable models is a major focus of ongoing research.

Papers

November 18, 2023

HungerGist: An Interpretable Predictive Model for Food Insecurity
Yongsu Ahn, Muheng Yan, Yu-Ru Lin, Zian Wang
Interpretable Model News Article Early Warning Food Security

November 17, 2023

November 3, 2023

Proto-lm: A Prototypical Network-Based Framework for Built-in Interpretability in Large Language Models
Sean Xie, Soroush Vosoughi, Saeed Hassanpour
Natural Language Processing Interpretable Model Prototypical Network Built in Interpretability Interpretable Embeddings

October 30, 2023

Bidirectional Captioning for Clinically Accurate and Interpretable Models
Keegan Quigley, Miriam Cha, Josh Barua, Geeticka Chauhan, Seth Berkowitz, Steven Horng, Polina Golland
Contrastive Learning Vision Language Image Captioning Interpretable Model Contrastive Pretraining

October 25, 2023

Interpretable time series neural representation for classification purposes
Etienne Le Naour, Ghislain Agoua, Nicolas Baskiotis, Vincent Guigue
Representation Learning Time Series Interpretable Model Neural Representation Symbolic Representation Classification Application

October 19, 2023

To grok or not to grok: Disentangling generalization and memorization on corrupted algorithmic datasets
Darshil Doshi, Aritra Das, Tianyu He, Andrey Gromov
Deep Learning Strong Generalization Regularization Model Interpretable Model Limited Memorization Robust Generalization Grokking Phenomenon

October 16, 2023

Interpretable Predictive Models to Understand Risk Factors for Maternal and Fetal Outcomes
Tomas M. Bosschieter, Zifei Xu, Hui Lan, Benjamin J. Lengerich, Harsha Nori, Ian Painter, Vivienne Souter, Rich Caruana
Predictive Model Interpretable Model Risk Factor Explainable Boosting Machine

October 5, 2023

October 4, 2023

Leveraging Model-based Trees as Interpretable Surrogate Models for Model Distillation
Julia Herbinger, Susanne Dandl, Fiona K. Ewald, Sofia Loibl, Giuseppe Casalicchio
Interpretable Model Model Distillation Interpretable Surrogate Model Model Tree

September 29, 2023

Efficient Interpretable Nonlinear Modeling for Multiple Time Series
Kevin Roy, Luis Miguel Lopez-Ramos, Baltasar Beferull-Lozano
Interpretable Model Nonlinear Model Time Series Prediction Vector Autoregressive

September 14, 2023

Doubly High-Dimensional Contextual Bandits: An Interpretable Model for Joint Assortment-Pricing
Junhui Cai, Ran Chen, Martin J. Wainwright, Linda Zhao
Contextual Bandit Interpretable Model Assortment Optimization Bandit Learning

August 29, 2023

Probabilistic Dataset Reconstruction from Interpretable Models
Julien Ferry, Ulrich Aïvodji, Sébastien Gambs, Marie-José Huguet, Mohamed Siala
Inherent Interpretability Interpretable Model Data Leakage Interpretable Regression

August 17, 2023

Interpretable Graph Neural Networks for Tabular Data
Amr Alkhatib, Sofiane Ennadir, Henrik Boström, Michalis Vazirgiannis
Graph Neural Network Representation Learning Tabular Data Interpretable Model Interpretable Graph Neural Network

August 9, 2023

Towards true discovery of the differential equations
Alexander Hvatov, Roman Titov
Scientific Discovery Interpretable Model Differential Equation Equation Discovery Differential Equation Discovery

Interpretable Model

Papers

HungerGist: An Interpretable Predictive Model for Food Insecurity

Flexible Model Interpretability through Natural Language Model Editing

Interpretable Modeling of Single-cell perturbation Responses to Novel Drugs Using Cycle Consistence Learning

Banach-Tarski Embeddings and Transformers

Cross-domain feature disentanglement for interpretable modeling of tumor microenvironment impact on drug response

Interpretable by Design: Wrapper Boxes Combine Neural Performance with Faithful Explanations

Proto-lm: A Prototypical Network-Based Framework for Built-in Interpretability in Large Language Models

Bidirectional Captioning for Clinically Accurate and Interpretable Models

Interpretable time series neural representation for classification purposes

To grok or not to grok: Disentangling generalization and memorization on corrupted algorithmic datasets

Interpretable Predictive Models to Understand Risk Factors for Maternal and Fetal Outcomes

Extreme sparsification of physics-augmented neural networks for interpretable model discovery in mechanics

The Blame Problem in Evaluating Local Explanations, and How to Tackle it

A Quantitatively Interpretable Model for Alzheimer's Disease Prediction Using Deep Counterfactuals

Leveraging Model-based Trees as Interpretable Surrogate Models for Model Distillation

Efficient Interpretable Nonlinear Modeling for Multiple Time Series

Doubly High-Dimensional Contextual Bandits: An Interpretable Model for Joint Assortment-Pricing

Probabilistic Dataset Reconstruction from Interpretable Models

Interpretable Graph Neural Networks for Tabular Data

Towards true discovery of the differential equations