Optimal quantum dataset for learning a unitary transformation

Abstract

Unitary transformations formulate the time evolution of quantum states. How to learn a unitary transformation efficiently is a fundamental problem in quantum machine learning. The most natural and leading strategy is to train a quantum machine learning model based on a quantum dataset. Although the presence of more training data results in better models, using too much data reduces the efficiency of training. In this work, we solve the problem on the minimum size of sufficient quantum datasets for learning a unitary transformation exactly, which reveals the power and limitation of quantum data. First, we prove that the minimum size of a dataset with pure states is $2^n$ for learning an $n$-qubit unitary transformation. To fully explore the capability of quantum data, we introduce a practical quantum dataset consisting of $n+1$ elementary tensor product states that are sufficient for exact training. The main idea is to simplify the structure utilizing decoupling, which leads to an exponential improvement in the size of the datasets with pure states. Furthermore, we show that the size of the quantum dataset with mixed states can be reduced to a constant, which yields an optimal quantum dataset for learning a unitary. We showcase the applications of our results in oracle compiling and Hamiltonian simulation. Notably, to accurately simulate a 3-qubit one-dimensional nearest-neighbor Heisenberg model, our circuit only uses $96$ elementary quantum gates, which is significantly less than $4080$ gates in the circuit constructed by the Trotter-Suzuki product formula.

11 pages including appendix, v2 added remarks and references, v3 is closed to the published version

References in corpus (32)

• Quantum Machine Learning
• Quantum machine learning in feature Hilbert spaces
• Quantum Convolutional Neural Networks
• Error mitigation for short-depth quantum circuits
• Quantum fingerprinting
• Parameterized quantum circuits as machine learning models
• Power of data in quantum machine learning
• Efficient Learning for Deep Quantum Neural Networks
• Toward the first quantum simulation with quantum speedup
• A rigorous and robust quantum speed-up in supervised machine learning
• Hybrid quantum-classical algorithms and quantum error mitigation
• Quantum autoencoders for efficient compression of quantum data
• Generalization in quantum machine learning from few training data
• Induced measures in the space of mixed quantum states
• Quantum-assisted quantum compiling
• Quantum channels and memory effects
• Quantum generalisation of feedforward neural networks
• Learning the quantum algorithm for state overlap
• Learning-based quantum error mitigation
• Noise Resilience of Variational Quantum Compiling
• Optimal quantum learning of a unitary transformation
• Machine learning of noise-resilient quantum circuits
• Expressibility of the alternating layered ansatz for quantum computation
• Using and reusing coherence to realize quantum processes
• Reformulation of the No-Free-Lunch Theorem for Entangled Data Sets
• Quasiprobability decompositions with reduced sampling overhead
• Quantum Channel Simulation and the Channel's Smooth Max-Information
• Robust quantum compilation and circuit optimisation via energy minimisation
• Noise-Assisted Quantum Autoencoder
• Entanglement cost and quantum channel simulation
• Physical Implementability of Linear Maps and Its Application in Error Mitigation
• Minimum number of input states required for quantum gate characterization

Cited by in corpus (6)

• Learning quantum states and unitaries of bounded gate complexity
• Quantum Phase Processing and its Applications in Estimating Phase and Entropies
• Generalization of Quantum Machine Learning Models Using Quantum Fisher Information Metric
• Quantum Channel Learning
• Assisted quantum simulation of open quantum systems
• Learning complexity gradually in quantum machine learning models