Latest Articles for Data Science

A novel method combines self-supervised and active learning for efficient activity recognition.

This method uses audio and visual data to summarize videos effectively.

A model that combines images and text for better conversations.

Learn how randomized smoothing strengthens machine learning against adversarial attacks.

A look at how gradient preconditioning enhances model training efficiency.

Explore how LLMs can structure sustainability information into Knowledge Graphs.

A new approach enhances KNN for better data classification.

A new method helps models learn from just one data example.

AI's self-supervised learning improves diabetic retinopathy detection with fewer labeled images.

Combining techniques to improve state estimation in carbon capture plants.

A new method enhances translation quality by generating extra training examples.

A new method enhances hyperparameter optimization through multi-objective approaches.

A novel framework improves recommendations and explanations through geometric user-item interactions.

A new method enhances GNNs, tackling oversmoothing and improving performance.

A new method improves topological data representation for machine learning.

Varta helps improve headline generation for Indic languages with 41 million articles.

A critical look at the reliability of the ROAR metric in feature importance.

New techniques help expand access to financial services for underserved communities.

Study reveals encoder-decoder models outperform traditional methods in label classification tasks.

Learn how Thompson Sampling improves choices in uncertain environments.

A look into generative models and their transformative role in data generation.

A new approach to enhance consistency in online product categorization.

DCR enhances AI interpretability while maintaining high accuracy.

VCSum provides a comprehensive dataset for effective meeting summarization in Chinese.

A new method enhances AI's ability to plan actions with specific requirements.

A new method aims to balance recommendations based on user preferences.

Improving estimates through variational data assimilation with Gauss-Newton method.

A new method analyzes COVID-19 rates and responses in England over time.

Learn how BPI improves imputation for datasets with missing values.

New knowledge base improves analogical reasoning in language models.

A software tool streamlines collection of leaderboard data from AI research papers.

Analyzing how LLMs predict user ratings effectively.

A new method enhances action recognition by focusing on timing and motion.

A new method combines active and continual learning for efficient model training.

A novel approach combines visual features and external knowledge for better question answering.

New method improves extraction of bacterial interactions from scientific literature.

XTab enhances machine learning for diverse tabular datasets, improving model flexibility.

This article discusses enhancing Gaussian Processes using neural networks and spherical inducing features.

Examining how deep learning models can produce unexpected outputs from small input changes.

Decker combines knowledge sources for improved commonsense verification.