A Differential Privacy Budget Allocation Algorithm Based on Out-of-Bag Estimation in Random Forest

Xin Li, Baodong Qin (), Yiyuan Luo and Dong Zheng
Additional contact information
Xin Li: School of Cyberspace Security, Xi’an University of Posts and Telecommunications, Xi’an 710121, China
Baodong Qin: School of Cyberspace Security, Xi’an University of Posts and Telecommunications, Xi’an 710121, China
Yiyuan Luo: School of Computer Science and Engineering, Huizhou University, Huizhou 516007, China
Dong Zheng: School of Cyberspace Security, Xi’an University of Posts and Telecommunications, Xi’an 710121, China

Mathematics, 2022, vol. 10, issue 22, 1-15

Abstract: The issue of how to improve the usability of data publishing under differential privacy has become one of the top questions in the field of machine learning privacy protection, and the key to solving this problem is to allocate a reasonable privacy protection budget. To solve this problem, we design a privacy budget allocation algorithm based on out-of-bag estimation in random forest. The algorithm firstly calculates the decision tree weights and feature weights by the out-of-bag data under differential privacy protection. Secondly, statistical methods are introduced to classify features into best feature set, pruned feature set, and removable feature set. Then, pruning is performed using the pruned feature set to avoid decision trees over-fitting when constructing an ϵ -differential privacy random forest. Finally, the privacy budget is allocated proportionally based on the decision tree weights and feature weights in the random forest. We conducted experimental comparisons with real data sets from Adult and Mushroom to demonstrate that this algorithm not only protects data security and privacy, but also improves model classification accuracy and data availability.

Keywords: differential privacy; machine learning; privacy protection; random forest; out-of-bag estimation (search for similar items in EconPapers)
JEL-codes: C (search for similar items in EconPapers)
Date: 2022
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2227-7390/10/22/4338/pdf (application/pdf)
https://www.mdpi.com/2227-7390/10/22/4338/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jmathe:v:10:y:2022:i:22:p:4338-:d:977472

Access Statistics for this article

Mathematics is currently edited by Ms. Emma He

More articles in Mathematics from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().