Guang Cheng

A look into the essential theory behind deep learning models.

Exploring how adversarial training improves model robustness through feature purification.

A look at how transformers excel with unstructured data in regression tasks.

A new method protects personal preferences while allowing useful rankings.

Exploring how neural networks can approximate functionals in data analysis.

TGDP improves data generation in scenarios with scarce training examples.

A new watermarking method aids in identifying synthetic tabular data.

Learn how backdoor attacks threaten machine learning systems and methods to defend against them.

New methods enhance optimization in complex scenarios using latent spaces and noise.

TimeAutoDiff offers new solutions for creating realistic synthetic time series data.

A new model improves the generation of synthetic tabular data for various applications.

Discover how synthetic data helps retailers protect customer privacy while gaining insights.

Exploring privacy risks in synthetic data and introducing the Data Plagiarism Index.

A three-step method for secure data sharing while protecting privacy.

Learn how watermarking protects against fake data in the digital world.

Proactive strategies using AI aim to fortify cloud security against emerging threats.