Enregistré dans:
Détails bibliographiques
Auteurs principaux: Zou, Tianyuan, Liu, Yang, Li, Peng, Xiong, Yufei, Zhang, Jianqing, Liu, Jingjing, Ye, Xiaozhou, Ouyang, Ye, Zhang, Ya-Qin
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2502.00245
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866909471903580160
author Zou, Tianyuan
Liu, Yang
Li, Peng
Xiong, Yufei
Zhang, Jianqing
Liu, Jingjing
Ye, Xiaozhou
Ouyang, Ye
Zhang, Ya-Qin
author_facet Zou, Tianyuan
Liu, Yang
Li, Peng
Xiong, Yufei
Zhang, Jianqing
Liu, Jingjing
Ye, Xiaozhou
Ouyang, Ye
Zhang, Ya-Qin
contents Substantial quantity and high quality are the golden rules of making a good training dataset with sample privacy protection equally important. Generating synthetic samples that resemble high-quality private data while ensuring Differential Privacy (DP), a formal privacy guarantee, promises scalability and practicality. However, existing methods relying on pre-trained models for data synthesis %that avoid fine-tuning large pre-trained generative models often struggle in data-deficient scenarios, suffering from limited sample size, inevitable generation noise and existing pre-trained model bias. To address these challenges, we propose a novel contrAstive private data Synthesis via Weighted multiple Pre-trained language models (PLM) framework, named as WASP. WASP utilizes limited private samples for more accurate private data distribution estimation via a Top-Q voting mechanism, and leverages low-quality synthetic samples for contrastive generation via collaboration among dynamically weighted multiple pre-trained models.Extensive experiments on 6 well-developed datasets with 6 open-source and 3 closed-source PLMs demonstrate the superiority of WASP in improving model performance over diverse downstream tasks. Code is available at https://anonymous.4open.science/r/WASP.
format Preprint
id arxiv_https___arxiv_org_abs_2502_00245
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Contrastive Private Data Synthesis via Weighted Multi-PLM Fusion
Zou, Tianyuan
Liu, Yang
Li, Peng
Xiong, Yufei
Zhang, Jianqing
Liu, Jingjing
Ye, Xiaozhou
Ouyang, Ye
Zhang, Ya-Qin
Machine Learning
Substantial quantity and high quality are the golden rules of making a good training dataset with sample privacy protection equally important. Generating synthetic samples that resemble high-quality private data while ensuring Differential Privacy (DP), a formal privacy guarantee, promises scalability and practicality. However, existing methods relying on pre-trained models for data synthesis %that avoid fine-tuning large pre-trained generative models often struggle in data-deficient scenarios, suffering from limited sample size, inevitable generation noise and existing pre-trained model bias. To address these challenges, we propose a novel contrAstive private data Synthesis via Weighted multiple Pre-trained language models (PLM) framework, named as WASP. WASP utilizes limited private samples for more accurate private data distribution estimation via a Top-Q voting mechanism, and leverages low-quality synthetic samples for contrastive generation via collaboration among dynamically weighted multiple pre-trained models.Extensive experiments on 6 well-developed datasets with 6 open-source and 3 closed-source PLMs demonstrate the superiority of WASP in improving model performance over diverse downstream tasks. Code is available at https://anonymous.4open.science/r/WASP.
title Contrastive Private Data Synthesis via Weighted Multi-PLM Fusion
topic Machine Learning
url https://arxiv.org/abs/2502.00245