Quantum Machine Learning

Quantum ML Foundations: Encoding, Variational Circuits, and the Parameter-Shift Rule

Quantum machine learning trains parameterized quantum circuits as models for classical data. This tutorial covers the three classical-to-quantum encoding strategies, the parameter-shift rule that makes gradient-based training possible, and a complete PennyLane example training a variational classifier on a real dataset.

intermediate · ~24 min · prereq: Tutorial 14: QAOA for Combinatorial Optimization

Quantum Kernels and Feature Maps

Quantum kernels sidestep variational training entirely: they embed data into a quantum Hilbert space via a fixed feature map and use the inner product as a kernel for a classical SVM. This tutorial builds the ZZ feature map from Havlíček et al. 2019, implements a quantum SVC in Qiskit, and explains the reproducing-kernel view that unifies the approach.

advanced · ~22 min · prereq: Tutorial 15: QML Foundations

Is QML Worth It? A Skeptic's Benchmark

Most published QML results test against toy baselines that serious classical ML would demolish. This tutorial runs a bake-off — variational QML, quantum kernels, XGBoost, and a small MLP — on real tabular data, surveys the 'dequantization' results that have taken quantum advantages back, and gives an honest recommendation on when to reach for QML vs not.

advanced · ~23 min · prereq: Tutorial 16: Quantum Kernels and Feature Maps

Tang Dequantization: How a Grad Student Took Back Years of Quantum-ML Speedups

In 2018, then-undergraduate Ewin Tang showed that a famous quantum-machine-learning algorithm (Kerenidis-Prakash recommendation systems) had a polynomial-time classical analog. Within two years, the same dequantization template had taken back claimed exponential speedups for several flagship QML algorithms. This tutorial covers the dequantization framework, what it does and doesn't take back, and the residual quantum-advantage candidates that survived.

advanced · ~17 min · prereq: Tutorial 17: Is QML Worth It? A Skeptic's Benchmark

Quantum Convolutional Neural Networks: Cong-Choi-Lukin and the Quantum-Data QML Story

Quantum convolutional neural networks (QCNNs) — Cong, Choi, and Lukin 2019 — are the QML architecture with the cleanest structural advantage on quantum-data inputs. They have a tree structure that avoids barren plateaus by construction, naturally implement renormalization-group-style coarse-graining, and are most useful for classifying quantum states (phases of matter, error syndromes, sensor outputs). This tutorial covers the architecture, the trainability proof, and the regimes where QCNNs actually win.

advanced · ~17 min · prereq: Tutorial 37: Barren Plateaus, Tutorial 41: Tang Dequantization

The Data-Loading Bottleneck: Why Quantum Machine Learning on Classical Data Rarely Delivers

The data-loading bottleneck is the structural reason most quantum machine learning on classical data does not deliver speedups. Loading $N$ classical bits into a quantum register typically costs $O(N)$ time — eating any algorithm's hoped-for $O(\log N)$ speedup. This tutorial covers the encoding options (amplitude, basis, angle), the QRAM construction and its open hardware question, and the regimes where the bottleneck is and is not binding.

advanced · ~16 min · prereq: Tutorial 41: Tang Dequantization

Quantum Generative Models: Born Machines, Quantum GANs, and the Sampling-Class Advantage

Quantum generative models — born machines, quantum generative adversarial networks (QGANs) — sample from probability distributions defined by quantum states. Unlike classification or regression QML, generative QML has a clean structural quantum-advantage argument: there exist distributions that quantum circuits can efficiently sample from but classical algorithms cannot. This tutorial covers born machines, QGANs, the sampling-class complexity argument, and the regimes where quantum generative models can plausibly beat classical alternatives.

advanced · ~16 min · prereq: Tutorial 41: Tang Dequantization, Tutorial 42: Quantum Convolutional Neural Networks