Model interpretability through the lens of computational complexity

P Barceló, M Monet, J Pérez… - Advances in neural …, 2020 - proceedings.neurips.cc
In spite of several claims stating that some models are more interpretable than others--eg,"
linear models are more interpretable than deep neural networks"--we still lack a principled …

On computing probabilistic abductive explanations

Y Izza, X Huang, A Ignatiev, N Narodytska… - International Journal of …, 2023 - Elsevier
The most widely studied explainable AI (XAI) approaches are unsound. This is the case with
well-known model-agnostic explanation approaches, and it is also the case with approaches …

A simple near-linear pseudopolynomial time randomized algorithm for subset sum

C **, H Wu - arxiv preprint arxiv:1807.11597, 2018 - arxiv.org
Given a multiset $ S $ of $ n $ positive integers and a target integer $ t $, the Subset Sum
problem asks to determine whether there exists a subset of $ S $ that sums up to $ t $. The …

On computing relevant features for explaining NBCs

Y Izza, J Marques-Silva - arxiv preprint arxiv:2207.04748, 2022 - arxiv.org
Despite the progress observed with model-agnostic explainable AI (XAI), it is the case that
model-agnostic XAI can produce incorrect explanations. One alternative are the so-called …

Strongly polynomial FPTASes for monotone dynamic programs

T Alon, N Halman - Algorithmica, 2022 - Springer
In this paper we introduce a framework for the automatic generation of Strongly Polynomial
Fully Polynomial Time Approximation Schemes (SFPTASes) for monotone dynamic …

Fully Polynomial Time Approximation Schemes for Robust Multistage Decision Making

N Halman, G Nannicini - INFORMS Journal on Computing, 2024 - pubsonline.informs.org
We design a framework to obtain Fully Polynomial Time Approximation Schemes (FPTASes)
for adjustable robust multistage decision making under the budgeted uncertainty sets …

Approximately Counting Knapsack Solutions in Subquadratic Time

W Feng, C ** - Proceedings of the 2025 Annual ACM-SIAM …, 2025 - SIAM
We revisit the classic# Knapsack problem, which asks to count the Boolean points (x 1, x
2,…, xn)∈{0, 1} n in a given half-space. This# P-complete problem is known to admit (1±∊) …

Synthetic Census Data Generation via Multidimensional Multiset Sum

C Dwork, K Greenewald, M Raghavan - arxiv preprint arxiv:2404.10095, 2024 - arxiv.org
The US Decennial Census provides valuable data for both research and policy purposes.
Census data are subject to a variety of disclosure avoidance techniques prior to release in …

Probability estimation via policy restrictions, convexification, and approximate sampling

A Chandra, M Tawarmalani - Mathematical Programming, 2022 - Springer
This paper develops various optimization techniques to estimate probability of events where
the optimal value of a convex program, satisfying certain structural assumptions, exceeds a …

Games with Weighted Multiple Objectives

O Kupferman, N Shenwald - … on Automated Technology for Verification and …, 2024 - Springer
Games with multiple objectives arise naturally in synthesis of reactive systems. We study
games with weighted multiple objectives. The winning objective in such games consists of a …