English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Despite significant effort, the quantum machine learning community has onlydemonstrated quantum learning advantages for artificial cryptography-inspireddatasets when dealing with classical data. In this paper we address thechallenge of finding learning problems where quantum learning algorithms canachieve a provable exponential speedup over classical learning algorithms. Wereflect on computational learning theory concepts related to this question anddiscuss how subtle differences in definitions can result in significantlydifferent requirements and tasks for the learner to meet and solve. We examineexisting learning problems with provable quantum speedups and find that theylargely rely on the classical hardness of evaluating the function thatgenerates the data, rather than identifying it. To address this, we present twonew learning separations where the classical difficulty primarily lies inidentifying the function generating the data. Furthermore, we explorecomputational hardness assumptions that can be leveraged to prove quantumspeedups in scenarios where data is quantum-generated, which implies likelyquantum advantages in a plethora of more natural settings (e.g., in condensedmatter and high energy physics). We also discuss the limitations of theclassical shadow paradigm in the context of learning separations, and howphysically-motivated settings such as characterizing phases of matter andHamiltonian learning fit in the computational learning framework.

Exponential separations between classical and quantum learners