Knowledge Tree
linear models
Also known as: linear model
Facts (10)
Sources
Track: Poster Session 3 - aistats 2026 virtual.aistats.org 4 facts
claimLeveraging optimization structure in linear models allows for significantly faster convergence rates compared to methods proposed in the context of deep learning.
claimRemi Khellaf, Aurélien Bellet, and Julie Josse derived the asymptotic variance of ATE estimators for linear models in the context of Randomized Controlled Trials (RCTs).
procedureAntonio Ribeiro, Thomas Schön, Dave Zachariah, and Francis Bach propose tailored optimization algorithms for the adversarial training of linear models, utilizing iterative ridge regression for regression problems and projected gradient descent for classification problems, based on extended variable reformulations of the original problem.
formulaAdversarial training for linear models can be formulated as a convex optimization problem.
A Survey on the Theory and Mechanism of Large Language Models arxiv.org Mar 12, 2026 3 facts
claimLinear models face two inherent difficulties: constant-size states cannot scale with sequence length, leading to information loss on long inputs, and compressed representations may fail if future patterns deviate from the prior encoded in the compression rule.
claimLinear models often gain computational efficiency at the expense of representational power, a phenomenon described by the 'no free lunch' principle.
claimFast weight programmers and online learners are a family of linear models obtained by applying different gradient-descent algorithms in online or streaming settings (Schmidhuber, 1992; Yang et al., 2024b; Liu et al., 2024a; Yang et al., 2024c).
Comprehensive framework for smart residential demand side ... nature.com Mar 22, 2025 1 fact
claimThe Residential Demand Side Management (RDSM) model has higher computational complexity compared to the linear model, despite being preferable to use.
A Copernican Approach to Brain Advancement: The Paradigm of ... frontiersin.org Apr 25, 2019 1 fact
claimThe homeostasis paradigm utilizes linear models, whereas the Paradigm of Allostatic Orchestration utilizes models based on complexity, criticality, and non-linear dynamics.
A Synergistic Workspace for Human Consciousness Revealed by ... elifesciences.org 1 fact
referenceThe article 'Macroscopic resting-state brain dynamics are best described by linear models' was published in Nature Biomedical Engineering in 2023.