Using Topological Data Analysis to classify Encrypted Bits

TL;DR

This paper introduces a novel approach using topological data analysis and persistent homology to classify encrypted bits, outperforming classical machine learning models and providing an effective dimensionality reduction technique.

Contribution

It demonstrates the application of topological data analysis for classifying encrypted data, a novel approach that enhances classification accuracy and reduces dimensionality.

Findings

Successful classification of encrypted bits using topological features

Outperforms classical machine learning models in this task

Provides an effective dimensionality reduction method

Abstract

We present a way to apply topological data analysis for classifying encrypted bits into distinct classes. Persistent homology is applied to generate topological features of a point cloud obtained from sets of encryptions. We see that this machine learning pipeline is able to classify our data successfully where classical models of machine learning fail to perform the task. We also see that this pipeline works as a dimensionality reduction method making this approach to classify encrypted data a realistic method to classify the given encryptioned bits.