Quantum Boosting

Suppose we have a weak learning algorithm for a Boolean-valued problem: produces hypotheses whose bias is small, only slightly better than random guessing (this could, for instance, be due to implementing on a noisy device), can we boost the performance of so that 's output is correct on of the inputs? Boosting is a technique that converts a weak and inaccurate machine learning algorithm into a strong accurate learning algorithm. The AdaBoost algorithm by Freund and Schapire (for which they were awarded the G\"odel prize in 2003) is one of the widely used boosting algorithms, with many applications in theory and practice. Suppose we have a -weak learner for a Boolean concept class that takes time , then the time complexity of AdaBoost scales as , where is the -dimension of . In this paper, we show how quantum techniques can improve the time complexity of classical AdaBoost. To this end, suppose we have a -weak quantum learner for a Boolean concept class that takes time , we introduce a quantum boosting algorithm whose complexity scales as thereby achieving a quadratic quantum improvement over classical AdaBoost in terms of .
View on arXiv