Research map / Bandits and online learning / Online convex optimization and games
Online Learning of Quantum States
Scott Aaronson, Xinyi Chen, Elad Hazan, Satyen Kale, Ashwin Nayak
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use RFTL with von Neumann entropy for online updates | Quantum-state hypotheses · Online quantum learning, convex L-Lipschitz loss | regret upper bound: 2L \sqrt{(2 \log 2) T n} | Improved |
| Use the Matrix Multiplicative Weights update instead of RFTL | Quantum-state hypotheses · Online quantum learning, convex L-Lipschitz loss | regret upper bound: 2L\sqrt{(\log 2)Tn} | Improved |
| Update only on rounds whose absolute loss exceeds 2ε/3 | RFTL or Matrix Multiplicative Weights · Online quantum learning, realizable feedback | mistakes with prediction error greater than ε: \operatorname{O}\!\left(\frac{n}{\varepsilon^2}\right) | Improved |
| Use postselection refinement instead of convex optimization | Quantum-state hypotheses · Online quantum learning, realizable feedback | mistakes with prediction error greater than ε: \operatorname{O}\!\!\left( \frac{n}{\varepsilon^{3}}\log\frac{n}{\varepsilon}\right) | Hurt |
| Use sequential fat-shattering dimension to obtain a non-constructive learner | Quantum-state hypotheses · Online quantum learning, non-realizable convex L-Lipschitz loss | regret upper bound: \operatorname{O} \left( L \sqrt{n T} \log^{3/2} T \right) | Hurt |
| Use binary measurement outcomes to form a stochastic gradient estimate | RFTL · Online quantum learning, realizable feedback | expected L2 regret: \operatorname{O}(\sqrt{Tn}\,) | Improved |
Closest papers by meaning
- Online Learning of Pure States is as Hard as Mixed States (NeurIPS 2025 · similarity 0.91)
- Limitations of measure-first protocols in quantum machine learning (ICML 2025 · similarity 0.87)
- Purest Quantum State Identification (NeurIPS 2025 · similarity 0.87)
- On the Trainability and Classical Simulability of Learning Matrix Product States Variationally (AAAI 2025 · similarity 0.84)
- Exponential Quantum Communication Advantage in Distributed Inference and Learning (NeurIPS 2024 · similarity 0.84)
- Toward the Goldilocks Blind Compression of Quantum States (NeurIPS 2026 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
