Tag
This paper investigates deep clustering methods on imbalanced tabular data and proposes two novel ensemble approaches that aggregate clustering assignments across embedding dimensions or via majority voting, outperforming individual methods on 16 datasets.
This paper studies how curriculum design and synthetic data diversity affect relational Prior-Data Fitted Network (PFN) pretraining, showing that progressive curricula can match or nearly match RDB-PFN's performance with far less synthetic data.
Introduces FairDiffuseVQVAE, a two-stage tabular diffusion model that achieves fairness at sampling time by conditioning on protected attributes, outperforming prior fair tabular generators on demographic parity and equalized odds.
This paper investigates context sampling for TabPFN on small tabular datasets, finding that context diversity and coverage are more important than distribution matching for accuracy, and that random sampling is effective.
TanML is an MIT-licensed automated model-validation toolkit for tabular machine-learning models, designed for regulated environments. The developers seek feedback on its features and reports.
This paper presents a distributed benchmark study on scaling laws for classical machine learning models on tabular data, showing that power-law fits hold for most model families and quantifying replicator-implementation variance across 127 student runs.
CLOE is a new semi-supervised anomaly detection method combining an autoencoder with a Christoffel Function-based detector, using a novel loss function to improve representation learning. It achieves state-of-the-art results on high-dimensional tabular data while maintaining simplicity.
This paper introduces a systematic framework for designing reproducible continual anomaly detection benchmarks from tabular datasets, and provides five benchmark scenarios from three cybersecurity datasets.
This paper introduces FALCON-Discover, a post-hoc framework for discovering regions where model predictions are confidently wrong, shifting calibration focus from aggregate metrics to sample-level dangerous failures.
Introduces K-IPO, a generate-then-select oversampling framework that preserves the original data's feature importance ranking (measured by Kendall's tau) during augmentation for imbalanced tabular data, showing improved preservation, explanation consistency, and predictive performance across 20 datasets.
This paper proposes a method that uses large language models to generate new features from tabular data via an evolutionary algorithm, demonstrating improved classification results across multiple datasets.
This study empirically compares Kolmogorov-Arnold Networks (KANs) and Multi-Layer Perceptrons (MLPs) on structured tabular classification tasks, finding that KANs statistically outperform MLPs but with higher computational cost.
MIT researcher Devavrat Shah's work on AI models for tabular time-series data led to the spinoff Ikigai Labs, recently acquired by Celonis, aiming to improve enterprise forecasting and decision-making.
Proposes a Task-Conditioned Synthetic Data Generation (TCSDG) algorithm combining a Bayesian Network generator with a transformer-based tabular foundation model to improve ML performance in agricultural prediction tasks, showing consistent improvements over benchmarks.
TabLoRA proposes a parameter-efficient neural ensemble method for large-scale tabular data by sharing a common backbone with predictor-specific low-rank adaptations, achieving competitive performance against GBDTs and deep learning baselines.
Proposes HDE-Net, a manifold-constrained deep neural network that uses hyperbolic space to better model rule-based structures in tabular data, achieving state-of-the-art performance on the TALENT-tiny-core benchmark while maintaining efficiency.
This paper evaluates state-of-the-art tabular models as encoders in image-tabular multimodal learning, addressing the challenge of using In-Context Learning models that require labels for embedding training and test instances.
This paper introduces SemiScope, an analysis tool designed to disentangle the effects of classifier tuning from joint SSL and classifier optimization in semi-supervised security classification. Results show that most performance gains attributed to joint optimization can be recovered by simply tuning the classifier and its decision threshold with Bayesian optimization.
This paper introduces MedKGTab, a knowledge-injected framework that uses biomedical knowledge graphs to expand cross-domain features in tabular medical data, addressing data scarcity by generating high-fidelity biomedical profiles.
TDGT is a web-based toolkit for synthetic tabular data generation that introduces the Adaptive Bayesian Mixture Synthesizer (ABMS) algorithm and a hybrid VAE-ABMS model, with GPU acceleration and comprehensive fidelity assessment.