Latest Articles for Data Science

Examining how model compression affects defenses against adversarial attacks.

MarginMatch improves model training with high-quality pseudo-labels.

New methods improve explainability and performance in multi-view learning for various fields.

Research highlights the limitations of transformer models with longer arithmetic inputs.

IGNNet aims to provide clear explanations for machine learning predictions.

Learn how optimal decision trees enhance decision-making in critical fields.

New framework improves learning from dynamic point cloud videos.

A new framework improves how recommendations align with user preferences.

Synthetic data generation enhances privacy while enabling valuable data analysis.

A new method enhances character recognition in comics by combining face and body features.

MarginMatch enhances learning from limited labeled data using model performance tracking.

This model simplifies explanations by combining multiple tasks into one.

Combining neural classifiers with Monte Carlo simulation improves data analysis methods.

Researchers improve object detection accuracy in traffic scenes using external knowledge.

This article examines weaknesses in test selection methods for deep learning systems.

A new method enhances motion prediction using auxiliary learning tasks.

Research highlights new ways to improve model defenses against adversarial attacks.

EDNIL enhances model robustness against changes in data contexts.

Machine learning helps predict water conditions in areas without monitoring stations.

Researchers fill gaps in star velocity data using advanced techniques.

A look at interval-closed sets and their toggling behavior.

Recent models enhance predictions for antibody interactions with key antigens.

Optimizing feature selection in clinical surveys through textual similarity analysis.

A fresh approach that combines ICL and code generation for improved predictions.

A look into graphs, their structures, and important concepts.

A new method enhances motion forecasting using self-supervised learning strategies.

Exploring machine learning models for better diagnosis of fungal infections.

Biquality learning tackles weak supervision and dataset shifts in machine learning effectively.

VALERIE22 dataset aids in developing safe automated driving through synthetic city images.

DEQDet introduces a fresh approach to object detection using a deep equilibrium model.

Improving continual learning through teacher adaptation and knowledge distillation.

This study investigates quantum computing techniques for improving satellite image acquisition scheduling.

This paper examines methods for assessing the relatedness of concepts using DBpedia.

Introducing Knowledge Bridge Learning to enhance knowledge transfer in deep learning.

A new method enhances query intent classification in online shopping environments.

Addressing real-world data issues with new learning methods.

NAPA-VQ offers a solution to retain knowledge in AI systems while learning new information.

A new approach to understanding graph-structured data and node relationships.

A new method for computing Shapley values in naive Bayes classifiers improves model interpretability.

A novel method improves the ability of Vision-Language Models to adapt to new tasks.