Quantum machine learning: Prospects and challenges
summary
Gives concrete algorithms and results: circuits to load vectors and compute dot products by loading then reversing the loader, a MNIST nearest-centroid experiment on an IonQ trapped-ion device achieves accuracy only a few points below classical (published in Nature Portfolio), orthogonal neural nets with a quantum forward pass and a quantum‑inspired backward pass trained in O(N^2) and run on IBM Pyramid circuits with medical images matching classical performance while preserving interpretability, a determinantal sampler with D^2 gates and D depth (vs classical D^3) and a GPU variant that yields a 7% precision lift for Itaú, logarithmic‑depth encoding of persistent homology, and quantum SVD tools for compound matrices.
AI-written summary · about
chapters (19)
- 0:00Intro
- 0:52Introduction to quantum computing
- 5:01Timeline of quantum computing
- 12:56Quantum machine learning
- 13:41Our approach to quantum ML
- 15:48First steps: Quantum ML
- 17:32Feasibility of quantum classifier
- 21:09Quantum orthogonal neural networks
- 21:40QNN for medical image classification
- 22:35Diversity is important
- 24:59Determinant sampling circuit
- 26:00Determinant sampling applications
- 26:19Quantum-enhanced ML for churn detection
- 27:42The big picture matters
- 30:45Topological data analysis computation
- 31:17Quantum topological data analysis
- 32:45Topological data analysis applications
- 34:25Compound matrix example 1 2 3 4
- 35:17Compound matrices
related talks
17mQuantum and Quantum-Inspired AI Solutions: Bridging Innovation and Industry Readiness
01h 06mPractical Quantum Machine Learning: Overcoming the Bottlenecks of Scalability and(…)
46mQuantum Machine Learning 101: An Introduction for Beginners with...
47mQuantum Machine Learning: Cutting Through the Buzzword BS
22mSecurity of Quantum Machine Learning Systems
29m10 Things You Should Know About Quantum Computing and ML
27mQuantum Machine Learning With TensorFlow Quantum
31mQuantum neural networks