Privacy-Preserving Boosting in the Local Setting

In machine learning, boosting is one of the most popular methods that designed to combine multiple base learners into a superior one. The well-known Boosted Decision Tree classifier has been widely adopted in data mining and pattern recognition. With emerging challenge privacy, personal images, browsing history, financial reports, which are held by individuals entities more likely contain sensitive information. privacy concern intensified when leaves hand owners used for further mining. Such issues demand learning algorithms should be privacy-aware. Recently, Local Differential Privacy proposed as an effective protection approach, allows perturb before any release. this paper, we propose distributed privacy-preserving algorithm can applied various types classifiers. By adopting LDP building block, leverages aggregation perturbed shares build learner, ensures well preserved participated owners. Our experiments demonstrate effectively boosts classifiers boosted maintain high utility.