A Quantum Bagging Algorithm with Unsupervised Base Learners for Label Corrupted Datasets

Neeshu Rathi, Sanjeev Kumar · 2025

Read A Quantum Bagging Algorithm with Unsupervised Base Learners for Label Corrupted Datasets by Neeshu Rathi, Sanjeev Kumar (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-07040.txt.

Description

Neeshu Rathi and Sanjeev Kumar propose a quantum bagging method whose base learners are unsupervised, in a 2025 paper. Quantum machine learning must cope with noisy hardware in the current era of intermediate-scale devices. Wrong labels in training data are a special problem, because supervised methods learn from those labels directly.

Their method clusters data with QMeans and uses that clustering as the base learner. Averaging over many resampled runs cuts prediction variance and makes the method harder to mislead with bad labels. Because QMeans needs no labels, it sidesteps much of the corruption. Sampling is done with quantum bootstrapping through QRAM, and the separate results are combined by majority vote. In broad tests on noisy classification and regression tasks, the method matched a classical KMeans version. It also held up better against label corruption than bagging built on supervised learners.

Catalog record

CREATORNeeshu Rathi, Sanjeev Kumar
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE46 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-07040

Browse the full library · About the Tower of Babel