Practical Quantum Machine Learning: Overcoming the Bottlenecks of Scalability and(…)
summary
Claims a method called QPMeL that maps data onto independent unit spheres, replacing entangling circuits with minimal rotations so hardware needs two gates per qubit. Training and metric optimization run off‑device so it pushes 100% classical optimization and avoids barren‑plateau entangling layers, cutting required qubits and depth exponentially. Paper reports scaling to full MNIST/Omniglot‑style tasks with only a handful of qubits, achieves state‑of‑the‑art multi‑class accuracy, and removes the need for thousands of measurement shots per gradient step.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:04Quantum systems constraints
- 2:11Tutorial overview
- 2:57Basics of ML classification
- 7:33Cross-entropy loss limits
- 13:56Deep metric learning
- 19:26Metric learning requirements
- 21:37Quantum geometry promise
- 27:47Cubit states fundamentals
- 33:34Quantum operators properties
- 41:04Measurement and fidelity
- 44:43Multi-qubit systems
- 46:32Naive quantum metric learning
- 53:23Distance calculation challenges
- 56:15Gradient calculation
- 1:00:01Scaling quantum metric learning
related talks
44mQuantum machine learning: Prospects and challenges
29m10 Things You Should Know About Quantum Computing and ML
15mAccelerQ: Accelerating Quantum Eigensolvers With Machine Learning on Quantum Simulators
22mSecurity of Quantum Machine Learning Systems
27mQuantum Machine Learning With TensorFlow Quantum
21mHybrid Quantum Computing for AI
19mAvenues for Quantum Error Correction and Mitigation
46mQuantum Machine Learning 101: An Introduction for Beginners with...