Dequantization and Quantum Advantage in Learning from Experiments - IPAM at UCLA

Explore the potential of quantum technology in revolutionizing data acquisition and processing for learning about the physical world in this 32-minute conference talk by Jarrod McClean from Google. Delve into the advantages of experimental setups that transduce data from physical systems to stable quantum memory and process it using quantum computers. Discover how quantum machines can learn from exponentially fewer experiments compared to conventional methods in tasks such as predicting physical system properties, performing quantum principal component analysis on noisy states, and learning approximate models of physical dynamics. Examine recent dequantization results in the context of quantum linear algebra and discuss the nuances related to dequantizing quantum algorithms. Gain insights into quantum machine learning, data-assisted problems, quantum memory, and quantum-enhanced experiments. Learn about Bell measurements as a feature in learning, experimental demonstrations of advantage, and unsupervised discovery and classification of processes. Understand the concepts of SWAPs and virtual distillation in quantum PCA, and explore the implications for future research in this field.

Intro
Quantum machine learning & data advantage
A motivating example for data assisted problems
The power of data in quantum machine learning
What kinds of problems are learnable from a little data?
Quantum memory and quantum-enhanced experiments
Quantum memory and quanturn-enhanced experiments
What's the simplest task we can have an advantage on?
The best possible conventional experiments
Summarizing the scale of the separation
Bell measurements as a feature in learning
Imagining and emulating a quantum data pipeline
Experimental demonstration of advantage
We've learned about states... how about processes?
Unsupervised discovery
Unsupervised classification of processes
SWAPs and virtual distillation to the quantum PCA
SWAP and virtual distillation to the quanturn PCA
Recal dequantization
Summary & Outlook