Tabular foundation models excel on tiny- to medium-sized IID data but are outperformed by traditional tree-based and deep learning models on non-IID, large, and high-dimensional datasets, based on evaluations across 11 models and 142 datasets in the new BeyondArena benchmark.
Breast Cancer
9 Pith papers cite this work, alongside 33 external citations. Polarity classification is still indexing.
9
Pith papers citing it
33
external citations · external index
citation-role summary
dataset 1
citation-polarity summary
roles
dataset 1polarities
use dataset 1representative citing papers
Leal learns cross-table alignment without shared features by treating lower training loss as evidence of correct row matching, with a cluster sampler to limit the candidate set.
The paper defines counterfactual explanations for Shapley values in data coalitions and proposes SV-Exp, a greedy algorithm that efficiently finds small data transfers to flip the value ranking.
A PPO-based classical agent (RLQ-Grad) replaces analytic QNN gradients, keeping update variance flat and gradient cost linear in parameters, but its practical gain is gated on the loss being non-flat.
Ensembles of UMAP projections, especially median-based consensus embeddings and input perturbation, preserve projection quality while revealing run-to-run variability and overfitting.
A systematic review of 53 papers on adversarial attacks for tabular machine learning finds the field fragmented, with efficacy over-emphasized and transferability, plausibility, and semantic preservation under-addressed.
Poisson-Burst and Delayed-Burst input encodings can lower membership-inference attack success in LIF spiking networks, with Poisson-Burst keeping accuracy close to rate coding.
FL-QDSNNs reports 94% accuracy on Iris using a dynamic-threshold Pauli-X spiking mechanism in federated quantum learning, but the evaluation lacks proper baselines and reproducibility details.
Columns of mixed numeric and categorical data are clustered by encoding categories as numbers, running factor analysis, and grouping columns by the factor that explains a majority of their variance.
citing papers explorer
-
Beyond IID: How General Are Tabular Foundation Models, Really?
Tabular foundation models excel on tiny- to medium-sized IID data but are outperformed by traditional tree-based and deep learning models on non-IID, large, and high-dimensional datasets, based on evaluations across 11 models and 142 datasets in the new BeyondArena benchmark.
-
Learning Relational Tabular Data without Shared Features
Leal learns cross-table alignment without shared features by treating lower training loss as evidence of correct row matching, with a cluster sampler to limit the candidate set.
-
Counterfactual Explanation of Shapley Value in Data Coalitions
The paper defines counterfactual explanations for Shapley values in data coalitions and proposes SV-Exp, a greedy algorithm that efficiently finds small data transfers to flip the value ranking.
-
Modeling quantum neural network gradient with reinforcement learning
A PPO-based classical agent (RLQ-Grad) replaces analytic QNN gradients, keeping update variance flat and gradient cost linear in parameters, but its practical gain is gated on the loss being non-flat.
-
Visualizing Uncertainty in Non-linear Projections with Ensembles
Ensembles of UMAP projections, especially median-based consensus embeddings and input perturbation, preserve projection quality while revealing run-to-run variability and overfitting.
-
Insights on Adversarial Attacks for Tabular Machine Learning via a Systematic Literature Review
A systematic review of 53 papers on adversarial attacks for tabular machine learning finds the field fragmented, with efficacy over-emphasized and transferability, plausibility, and semantic preservation under-addressed.
-
Izhikevich-Inspired Temporal Dynamics for Enhancing Privacy, Efficiency, and Transferability in Spiking Neural Networks
Poisson-Burst and Delayed-Burst input encodings can lower membership-inference attack success in LIF spiking networks, with Poisson-Burst keeping accuracy close to rate coding.
-
FL-QDSNNs: Federated Learning with Quantum Dynamic Spiking Neural Networks
FL-QDSNNs reports 94% accuracy on Iris using a dynamic-threshold Pauli-X spiking mechanism in federated quantum learning, but the evaluation lacks proper baselines and reproducibility details.
-
New Approach to Clustering Random Attributes
Columns of mixed numeric and categorical data are clustered by encoding categories as numbers, running factor analysis, and grouping columns by the factor that explains a majority of their variance.