Latest Articles for Data Science

A framework to improve user recommendations by blending familiarity with new options.

A new method combines decision trees and transformers for better decision-making.

A new method uses internal states for better accuracy in LLMs.

Triplet Graph Transformer enhances prediction of molecular properties through direct atom communication.

Learn how new techniques improve the efficiency of large machine learning models.

New methods enhance sample efficiency and speed in reinforcement learning.

This study investigates adversarial attacks on deep learning, focusing on CNN performance.

Text2Data improves data generation using unlabeled and labeled sources effectively.

Research highlights medication data's role in improving ICD coding accuracy.

A new method improves model accuracy by addressing noisy labels in federated learning.

Examining how GFlowNets generalize to untested areas and their application potential.

New methods enhance the clarity of machine learning predictions.

This article examines how neural networks manage resources for complex tasks.

This study improves GNN performance through effective graph explanations.

A study explores new methods for deep learning models to adapt to changing data.

A new method enhances search engines by using images for query suggestions.

A method to improve dense retriever selection using LLMs without requiring labels.

BOWLL sets a new standard for adaptive learning systems.

A novel approach merging traditional models with machine learning improves infection predictions.

RC-GNN enhances GNN interpretability and predictive accuracy through innovative methods.

Study highlights issues affecting fairness in AI explanations across different groups.

This study explores methods to improve classifier performance on imbalanced datasets.

A novel method predicts behavior of control-affine nonlinear systems effectively.

Introducing ApiQ for improved fine-tuning and quantization of large language models.

Longer instructions enhance language model performance and reduce complexity.

New method enhances reliability of model predictions in real-world applications.

A new approach to integrate diverse cultural insights into language models.

Introducing a dataset to test AI's understanding of physical interactions with soft bodies and liquids.

Introducing CAP to improve fairness and efficiency in machine learning models.

Quilt improves multi-class classification on today’s quantum machines.

A new method offers clearer insights into GNN predictions without extra training.

This research combines Fourier analysis with diffusion models for better time series generation.

A method combining PCA and statistical depth improves outlier detection in data analysis.

Exploring issues in creating decision-making AI models and solutions.

A new method for faster and better machine learning program searches.

A study on how parameter adjustments can improve algorithm performance.

Examining the link between dataset compression and adversarial robustness in deep learning.

New techniques improve Bayesian inference in complex, high-dimensional systems.

A2GNN model enhances performance in adapting graph knowledge across different domains.

New method optimizes sampling by combining it with optimization techniques.