We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

quant-ph

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Quantum Physics

Title: Quantum Cross Entropy and Maximum Likelihood Principle

Abstract: Quantum machine learning is an emerging field at the intersection of machine learning and quantum computing. Classical cross entropy plays a central role in machine learning. We define its quantum generalization, the quantum cross entropy, prove its lower bounds, and investigate its relation to quantum fidelity. In the classical case, minimizing cross entropy is equivalent to maximizing likelihood. In the quantum case, when the quantum cross entropy is constructed from quantum data undisturbed by quantum measurements, this relation holds. Classical cross entropy is equal to negative log-likelihood. When we obtain quantum cross entropy through empirical density matrix based on measurement outcomes, the quantum cross entropy is lower-bounded by negative log-likelihood. These two different scenarios illustrate the information loss when making quantum measurements. We conclude that to achieve the goal of full quantum machine learning, it is crucial to utilize the deferred measurement principle.
Comments: 8 pages, 1 figure. Section added. Submission to PRX
Subjects: Quantum Physics (quant-ph); Information Theory (cs.IT); Machine Learning (cs.LG); High Energy Physics - Theory (hep-th); Machine Learning (stat.ML)
Cite as: arXiv:2102.11887 [quant-ph]
  (or arXiv:2102.11887v2 [quant-ph] for this version)

Submission history

From: Zhou Shangnan [view email]
[v1] Tue, 23 Feb 2021 19:00:06 GMT (14kb)
[v2] Thu, 26 Aug 2021 17:19:20 GMT (68kb,D)

Link back to: arXiv, form interface, contact.