An Assessment of the Application of Private Aggregation of Ensemble Models to Sensible Data
This paper explores the use of Private Aggregation of Teacher Ensembles (PATE) in a setting where students have their own private data that cannot be revealed as is to the ensemble. We propose a privacy model that introduces a local differentially private mechanism to protect student data. We implem...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-09-01
|
Series: | Machine Learning and Knowledge Extraction |
Subjects: | |
Online Access: | https://www.mdpi.com/2504-4990/3/4/39 |
_version_ | 1797502882348531712 |
---|---|
author | Sergio Yovine Franz Mayr Sebastián Sosa Ramiro Visca |
author_facet | Sergio Yovine Franz Mayr Sebastián Sosa Ramiro Visca |
author_sort | Sergio Yovine |
collection | DOAJ |
description | This paper explores the use of Private Aggregation of Teacher Ensembles (PATE) in a setting where students have their own private data that cannot be revealed as is to the ensemble. We propose a privacy model that introduces a local differentially private mechanism to protect student data. We implemented and analyzed it in case studies from security and health domains, and the result of the experiment was twofold. First, this model does not significantly affecs predictive capabilities, and second, it unveiled interesting issues with the so-called data dependency privacy loss metric, namely, high variance and values. |
first_indexed | 2024-03-10T03:42:30Z |
format | Article |
id | doaj.art-eca18f06822b44b4802590d3e25db122 |
institution | Directory Open Access Journal |
issn | 2504-4990 |
language | English |
last_indexed | 2024-03-10T03:42:30Z |
publishDate | 2021-09-01 |
publisher | MDPI AG |
record_format | Article |
series | Machine Learning and Knowledge Extraction |
spelling | doaj.art-eca18f06822b44b4802590d3e25db1222023-11-23T09:17:33ZengMDPI AGMachine Learning and Knowledge Extraction2504-49902021-09-013478880110.3390/make3040039An Assessment of the Application of Private Aggregation of Ensemble Models to Sensible DataSergio Yovine0Franz Mayr1Sebastián Sosa2Ramiro Visca3Facultad de Ingeniería, Universidad ORT Uruguay, Montevideo 11100, UruguayFacultad de Ingeniería, Universidad ORT Uruguay, Montevideo 11100, UruguayFacultad de Ingeniería, Universidad ORT Uruguay, Montevideo 11100, UruguayFacultad de Ingeniería, Universidad ORT Uruguay, Montevideo 11100, UruguayThis paper explores the use of Private Aggregation of Teacher Ensembles (PATE) in a setting where students have their own private data that cannot be revealed as is to the ensemble. We propose a privacy model that introduces a local differentially private mechanism to protect student data. We implemented and analyzed it in case studies from security and health domains, and the result of the experiment was twofold. First, this model does not significantly affecs predictive capabilities, and second, it unveiled interesting issues with the so-called data dependency privacy loss metric, namely, high variance and values.https://www.mdpi.com/2504-4990/3/4/39machine learningdifferential privacyprivate aggregation of teacher ensemble |
spellingShingle | Sergio Yovine Franz Mayr Sebastián Sosa Ramiro Visca An Assessment of the Application of Private Aggregation of Ensemble Models to Sensible Data Machine Learning and Knowledge Extraction machine learning differential privacy private aggregation of teacher ensemble |
title | An Assessment of the Application of Private Aggregation of Ensemble Models to Sensible Data |
title_full | An Assessment of the Application of Private Aggregation of Ensemble Models to Sensible Data |
title_fullStr | An Assessment of the Application of Private Aggregation of Ensemble Models to Sensible Data |
title_full_unstemmed | An Assessment of the Application of Private Aggregation of Ensemble Models to Sensible Data |
title_short | An Assessment of the Application of Private Aggregation of Ensemble Models to Sensible Data |
title_sort | assessment of the application of private aggregation of ensemble models to sensible data |
topic | machine learning differential privacy private aggregation of teacher ensemble |
url | https://www.mdpi.com/2504-4990/3/4/39 |
work_keys_str_mv | AT sergioyovine anassessmentoftheapplicationofprivateaggregationofensemblemodelstosensibledata AT franzmayr anassessmentoftheapplicationofprivateaggregationofensemblemodelstosensibledata AT sebastiansosa anassessmentoftheapplicationofprivateaggregationofensemblemodelstosensibledata AT ramirovisca anassessmentoftheapplicationofprivateaggregationofensemblemodelstosensibledata AT sergioyovine assessmentoftheapplicationofprivateaggregationofensemblemodelstosensibledata AT franzmayr assessmentoftheapplicationofprivateaggregationofensemblemodelstosensibledata AT sebastiansosa assessmentoftheapplicationofprivateaggregationofensemblemodelstosensibledata AT ramirovisca assessmentoftheapplicationofprivateaggregationofensemblemodelstosensibledata |