no code implementations • 6 Feb 2025 • Alexander Asemota, Giles Hooker
Prior work in fairness inference has largely focused on inferring the fairness properties of a given predictive algorithm.
no code implementations • 12 Nov 2024 • Yunzhe Zhou, Giles Hooker
Did they Reproduce?
no code implementations • 4 Nov 2024 • Xiaohan Wang, Yunzhe Zhou, Giles Hooker
Variable importance is one of the most widely used measures for interpreting machine learning with significant interest from both statistics and machine learning communities.
no code implementations • 3 Jul 2024 • Kevin Tan, Giles Hooker, Edward L. Ionides
Automatic differentiation (AD) has driven recent advances in machine learning, including deep neural networks and Hamiltonian Markov Chain Monte Carlo methods.
1 code implementation • 14 Jun 2024 • Facundo Sapienza, Jordi Bolibar, Frank Schäfer, Brian Groenke, Avik Pal, Victor Boussange, Patrick Heimbach, Giles Hooker, Fernando Pérez, Per-Olof Persson, Christopher Rackauckas
Many scientific models are based on differential equations, where differentiable programming plays a crucial role in calculating model sensitivities, inverting model parameters, and training hybrid models that combine differential equations with data-driven approaches.
no code implementations • 29 Apr 2024 • Xi Xin, Giles Hooker, Fei Huang
The adoption of artificial intelligence (AI) across industries has led to the widespread use of complex black-box models and interpretation tools for decision making.
no code implementations • 29 Feb 2024 • Alexander Asemota, Giles Hooker
Counterfactual explanations are a common approach to providing recourse to data subjects.
2 code implementations • 28 Jan 2024 • Jeremy Goldwasser, Giles Hooker
Feature importance scores are ubiquitous tools for understanding the predictions of machine learning models.
1 code implementation • 11 Oct 2023 • Jeremy Goldwasser, Giles Hooker
Shapley values are among the most popular tools for explaining predictions of blackbox machine learning models.
1 code implementation • 22 Nov 2022 • Yunzhe Zhou, Peiru Xu, Giles Hooker
Model distillation has been a popular method for producing interpretable machine learning.
1 code implementation • 31 Aug 2022 • Indrayudh Ghosal, Yunzhe Zhou, Giles Hooker
In this paper we extend the Infinitesimal Jackknife to estimate the covariance between any two models.
2 code implementations • 15 Jun 2021 • Zhengze Zhou, Giles Hooker, Fei Wang
An increasing number of machine learning models have been deployed in domains with high stakes such as finance and healthcare.
no code implementations • 24 Feb 2021 • Indrayudh Ghosal, Giles Hooker
We use these residuals and some corresponding weights to fit a base random forest and then repeat the same to obtain a boost random forest.
no code implementations • 23 Feb 2021 • Lucas Mentch, Giles Hooker
In 2001, Leo Breiman wrote of a divide between "data modeling" and "algorithmic modeling" cultures.
1 code implementation • 2 Dec 2019 • Zhengze Zhou, Lucas Mentch, Giles Hooker
This paper develops a general framework for analyzing asymptotics of $V$-statistics.
1 code implementation • 12 Nov 2019 • Benjamin Lengerich, Sarah Tan, Chun-Hao Chang, Giles Hooker, Rich Caruana
Models which estimate main effects of individual variables alongside interaction effects have an identifiability challenge: effects can be freely moved between main effects and interaction effects without changing the model prediction.
1 code implementation • 1 May 2019 • Giles Hooker, Lucas Mentch, Siyu Zhou
This paper reviews and advocates against the use of permute-and-predict (PaP) methods for interpreting black box functions.
1 code implementation • 1 Apr 2019 • Yichen Zhou, Giles Hooker
This paper investigates the integration of gradient boosted decision trees and varying coefficient models.
Methodology
1 code implementation • 12 Mar 2019 • Zhengze Zhou, Giles Hooker
We propose a modification that corrects for split-improvement variable importance measures in Random Forests and other tree-based methods.
no code implementations • 22 Aug 2018 • Yichen Zhou, Zhengze Zhou, Giles Hooker
Here, we consider the use of regression trees as a student model, in which nodes of the tree can be used as `explanations' for particular predictions, and the whole structure of the tree can be used as a global representation of the resulting function.
1 code implementation • 26 Jun 2018 • Yichen Zhou, Giles Hooker
This paper examines a novel gradient boosting framework for regression.
1 code implementation • 21 Mar 2018 • Indrayudh Ghosal, Giles Hooker
In this paper we propose using the principle of boosting to reduce the bias of a random forest prediction in the regression setting.
1 code implementation • ICLR 2019 • Sarah Tan, Giles Hooker, Paul Koch, Albert Gordo, Rich Caruana
In this paper, we study global additive explanations for non-additive models, focusing on four explanation methods: partial dependence, Shapley explanations adapted to a global setting, distilled additive explanations, and gradient-based explanations.
no code implementations • 19 Nov 2017 • Skyler Seto, Sarah Tan, Giles Hooker, Martin T. Wells
Non-negative matrix factorization (NMF) is a technique for finding latent representations of data.
1 code implementation • 17 Oct 2017 • Sarah Tan, Rich Caruana, Giles Hooker, Yin Lou
We compare the student model trained with distillation to a second un-distilled transparent model trained on ground-truth outcomes, and use differences between the two models to gain insight into the black-box model.
no code implementations • 15 Apr 2017 • Giles Hooker, Cliff Hooker
Among the consequences just might be the triumph of anti-realism over realism.
2 code implementations • 22 Nov 2016 • Sarah Tan, Matvey Soloviev, Giles Hooker, Martin T. Wells
Ensembles of decision trees perform well on many problems, but are not interpretable.
no code implementations • 1 Jun 2015 • Giles Hooker, Lucas Mentch
This paper examines the use of a residual bootstrap for bias correction in machine learning regression methods.
no code implementations • 7 Jun 2014 • Lucas Mentch, Giles Hooker
While statistical learning methods have proved powerful tools for predictive modeling, the black-box nature of the models they produce can severely limit their interpretability and the ability to conduct formal inference.
no code implementations • 25 Apr 2014 • Lucas Mentch, Giles Hooker
Instead of aggregating full bootstrap samples, we consider predicting by averaging over trees built on subsamples of the training set and demonstrate that the resulting estimator takes the form of a U-statistic.
1 code implementation • 13 Oct 2012 • Giles Hooker, Kevin K. Lin, Bruce Rogers
We propose to accomplish this with a two-step strategy: when the full state vector of the diffusion process is observable continuously, we formulate this as an optimal control problem and apply numerical techniques from stochastic optimal control to solve it.
Methodology