Privacy-Preserving Chaotic Extreme Learning Machine with Fully Homomorphic Encryption

TL;DR

This paper introduces a novel privacy-preserving chaotic extreme learning machine utilizing fully homomorphic encryption, which maintains data confidentiality while achieving comparable or better performance than traditional models.

Contribution

It proposes a new chaotic extreme learning machine with encrypted weights and biases generated via logistic map, enhancing privacy in machine learning models.

Findings

Performs better or similar to traditional extreme learning machine on most datasets

Uses fully homomorphic encryption for data security

Introduces chaotic weight generation with logistic map

Abstract

The Machine Learning and Deep Learning Models require a lot of data for the training process, and in some scenarios, there might be some sensitive data, such as customer information involved, which the organizations might be hesitant to outsource for model building. Some of the privacy-preserving techniques such as Differential Privacy, Homomorphic Encryption, and Secure Multi-Party Computation can be integrated with different Machine Learning and Deep Learning algorithms to provide security to the data as well as the model. In this paper, we propose a Chaotic Extreme Learning Machine and its encrypted form using Fully Homomorphic Encryption where the weights and biases are generated using a logistic map instead of uniform distribution. Our proposed method has performed either better or similar to the Traditional Extreme Learning Machine on most of the datasets.