Latest Articles for Data Science

Exploring the security challenges of prompt engineering with LLMs.

A new Bayesian approach improves methods for learning DAG structures from data.

This study examines how language models learn and store information during training.

Learn how MTFA reduces data dimensions for clearer insights.

This article explores how randomness affects learning with limited labeled data.

Algebra introduces new methods for handling complex data in machine learning.

An overview of SKI-CL framework for improved time series predictions.

Introducing a novel approach to rank language models without needing reference answers.

Enhancing computer understanding of images and text through advanced training techniques.

A new approach enhances predictions of material properties using self-supervised learning.

A new model improves 3D understanding using varied datasets for better performance.

A new method improves estimation for minimizing losses in nonlinear models.

Enhance communication with LLMs by understanding errors and using clear prompts.

Organizing training data improves language model performance significantly.

This study presents a neural network designed to understand periodic systems.

This study examines biases in masked language models and their implications.

This article examines methods to enhance confidence levels in language models.

Introducing Kuaiji, an advanced model tailored for accounting professionals.

MSTAR improves time series classification using automated architecture and frequency resolution.

A look into the significance of the Local Learning Coefficient in machine learning models.

A new system enhances product recommendations using user data and expert insights.

New methods improve speed and accuracy in atmospheric chemistry modeling.

A new method improves graph continual learning by enhancing diversity in replay buffers.

A new approach uses neural networks to speed up eigenvalue and eigenvector finding.

A new method to improve machine learning models affected by poor data.

New methods improve how models learn from data for better predictions.

A new method enhances energy consumption predictions using physical principles.

New dataset enhances machines' ability to process image-related questions.

This article presents a framework for studying interactions between multiple populations over time.

A method to fix backdoor issues in foundation models without losing functionality.

Introducing D-SOBA, a new algorithm for decentralized bilevel optimization.

Learn how knowledge distillation improves smaller models using insights from larger ones.

Exploring data augmentation techniques and their impact on NLP models.

A new method enhances model training for better future predictions.

A new method helps robots learn tasks using simulated data.

A method to enhance machine learning using unlabeled data across various domains.

A look at M2MeT 2.0 and its impact on meeting transcription.

This article examines the impact of LLMs across various sectors.

FUTE offers a new approach to task embeddings across multiple models.

New coding method enhances data compression without quantization.