New Research by TSYS School's Peker and Raj Uses Homomorphic Encryption to Perform Statistical Analysis

Homomorphic encryption allows computations on encrypted data without revealing it to anyone other than an owner or an authorized collector. When combined with other techniques, homomorphic encryption offers an ideal solution for ensuring statistical confidentiality. TFHE (Fast Fully Homomorphic Encryption over the Torus) is a fully homomorphic encryption scheme that supports efficient homomorphic operations on Booleans and integers. A new study by TSYS School computer scientist Yesem Kurt Peker and her student Rahul Raj uses Zama’s Concrete compiler to explore the application of TFHE for performing statistical analysis on encrypted data, thereby demonstrating its viability for ensuring statistical confidentiality. The two researchers provide implementations of traditional algorithms for basic statistical computations on encrypted datasets, including the five-number summary, mean, variance, and mode, and record the time required for each operation. The results show that basic tasks like mean and min/max work well for small datasets while keeping data encrypted. However, more complex tasks like median and variance slow down dramatically as datasets get larger. This work reinforces the theoretical promise of Fully Homomorphic Encryption (FHE) for statistical analysis and highlights the need for substantial optimizations to make it viable for real-world applications.