| name | quantum-cv-learning-theory |
| description | Quantum learning theory for continuous-variable (bosonic) systems. Covers sample complexity analysis for learning non-Gaussian and Gaussian states, trace distance bounds via covariance matrices, Gaussian state testing, and efficient Gaussian process learning protocols. Use when: quantum learning theory, continuous variable, bosonic systems, Gaussian states, quantum tomography, sample complexity quantum, CV quantum, quantum state learning, quantum optical systems, bosonic quantum learning.
|
Quantum CV Learning Theory
Core Framework
Quantum learning theory for continuous-variable (CV) systems addresses: how efficiently can classical information be extracted from bosonic quantum systems? Covers sample complexity, state discrimination, and process learning.
Key Questions Addressed
1. Sample Complexity of Non-Gaussian States
- Minimum copies required to learn a non-Gaussian CV state
- Energy-constrained learning bounds
- Scaling with non-Gaussianity measures
2. Sample Complexity of Gaussian States
- Optimal copy requirements for Gaussian state learning
- Dependence on number of modes
- Efficient tomography protocols
3. Gaussianity Testing
- Test whether a state is Gaussian vs. far from Gaussian set
- Sample-efficient hypothesis testing
- Non-Gaussianity as a resource quantifier
4. Gaussian Process Learning
- Efficient learning of Gaussian processes (channels)
- Process tomography for CV operations
- Channel discrimination protocols
Trace Distance Bounds
Central tool: bounds on trace distance between CV states via their covariance matrices:
$$D(\rho, \sigma) \leq f(\Gamma_\rho, \Gamma_\sigma)$$
where $\Gamma$ denotes covariance matrices. These bounds enable efficient state comparison without full tomography.
Practical Applications
- Quantum sensing: sample-efficient characterization of bosonic probes
- Quantum communication: learning channel properties for optimal encoding
- Quantum computing: CV state preparation verification
- Quantum optics: experimental state characterization
Sample Complexity Scaling
| Task | Sample Complexity | Constraint |
|---|
| Gaussian state learning | O(n²) | n modes |
| Non-Gaussian state learning | Depends on non-Gaussianity | Energy bounded |
| Gaussianity testing | O(n/ε²) | ε accuracy |
| Gaussian process learning | O(n²) | per mode |
References
- arXiv: 2605.08082 - "Advances in quantum learning theory with bosonic systems" by Francesco Anna Mele
