Non-linear classification capability of quantum neural networks due to emergent quantum metastability

Abstract

The power and expressivity of deep classical neural networks can be attributed to non-linear input-output relations. Such non-linearities are at the heart of many computational tasks, such as data classification and pattern recognition. Quantum neural networks, on the other hand, are necessarily linear as they process information via unitary operations. Here we show that effective non-linearities can be implemented in these platforms by exploiting the relationship between information processing and many-body quantum dynamics. The crucial point is that quantum many-body systems can show emergent collective behavior in the vicinity of phase transitions, which leads to an effectively non-linear dynamics in the thermodynamic limit. In the context of quantum neural networks, which are necessarily finite, this translates into metastability with transient non-ergodic behavior. By using a quantum neural network whose architecture is inspired by dissipative many-body quantum spin models, we show that this mechanism indeed allows to realize non-linear data classification, despite the underlying dynamics being local and linear. Our proof-of-principle study may pave the way for the systematic construction of quantum neural networks with emergent non-linear properties.

arXiv comment19 pages, 6 figures

References in corpus (43)

• Quantum Computing in the NISQ era and beyond
• Quantum Machine Learning
• Variational Quantum Algorithms
• Barren plateaus in quantum neural network training landscapes
• A Practical Introduction to Tensor Networks: Matrix Product States and Projected Entangled Pair States
• Efficient Learning for Deep Quantum Neural Networks
• Challenges and Opportunities in Quantum Machine Learning
• Time-evolution methods for matrix-product states
• The quest for a Quantum Neural Network
• Quantum generative adversarial networks
• Continuous-variable quantum neural networks
• Quantum collision models: open system dynamics from repeated interactions
• Towards a theory of metastability in open quantum dynamics
• Barren Plateaus in Variational Quantum Computing
• Is quantum advantage the right goal for quantum machine learning?
• Variational principle for steady states of dissipative quantum many-body systems
• Universal non-equilibrium properties of dissipative Rydberg gases
• Trainability of Dissipative Perceptron-Based Quantum Neural Networks
• Dynamical phases and intermittency of the dissipative quantum Ising model
• Opportunities in Quantum Reservoir Computing and Extreme Learning Machines
• Quantum autoencoders to denoise quantum data
• Quantum computing models for artificial neural networks
• Composite quantum collision models
• Collision models can efficiently simulate any multipartite Markovian quantum dynamics
• Multistability of Driven-Dissipative Quantum Spins
• Collision models in quantum optics
• Bistability vs. Metastability in Driven Dissipative Rydberg Gases
• Multicritical behavior in dissipative Ising models
• Phase diagram of the dissipative quantum Ising model on a square lattice
• Metastability in an open quantum Ising model
• Open quantum generalisation of Hopfield neural networks
• Supervised Hebbian Learning
• Theory of classical metastability in open quantum systems
• Non-equilibrium phase transitions in $(1+1)$-dimensional quantum cellular automata with controllable quantum correlations
• Non-equilibrium absorbing state phase transitions in discrete-time quantum dynamics
• A brief journey through collision models for multipartite open quantum dynamics
• Phase diagram of quantum generalized Potts-Hopfield neural networks
• Quantum and classical temporal correlations in $(1+1)D$ Quantum Cellular Automata
• Using (1 + 1)D Quantum Cellular Automata for Exploring Collective Effects in Large Scale Quantum Neural Networks
• Numerical Simulation of Critical Quantum Dynamics without Finite Size Effects
• Optimal storage capacity of quantum Hopfield neural networks
• Asynchronism and nonequilibrium phase transitions in $(1+1)$D quantum cellular automata
• Dissipative quantum many-body dynamics in (1+1)D quantum cellular automata and quantum neural networks

Cited by in corpus (2)

• Relaxation Control of Open Quantum Systems
• Nonequilibrium phases and quantum correlations in synthetic transport models