The definitive review of the NISQ-era workhorse: ansatz design, optimisers, barren plateaus and applications, with the open problems stated clearly.
Variational algorithms (VQE & QAOA)
Hybrid loops where a shallow parameterised quantum circuit produces expectation values and a classical optimiser tunes the parameters — the dominant paradigm for near-term hardware.
Why it matters
VQE and QAOA are what most NISQ-era experiments actually run. Understanding them includes understanding their problems: barren plateaus, measurement overhead, and the absence of proven speedups.
After this you will be able to
- Implement VQE for a small molecule and QAOA for MaxCut
- Explain barren plateaus and strategies against them
- Assess whether a variational result beats classical methods
3 best places to start
Hand-picked and ordered. If you only have time for one, take the first.
The original QAOA paper, and short. It launched an enormous literature on near-term optimisation, most of which is still arguing about whether the speedup exists.
Several hundred runnable, paper-linked notebooks on variational algorithms, quantum chemistry and QML. The closest thing the field has to a "reproduce this result" library.
6 more resources
The library for compiling fermionic and electronic-structure problems into qubit Hamiltonians. Where quantum chemistry meets quantum circuits in practice.
Learn by writing code in the browser, exercise by exercise, from single qubits up to variational algorithms. The best option if you retain things by typing rather than reading.
Coding tutorials, seminar talks and the complete recordings of the Qiskit Global Summer Schools — which are effectively free graduate-level lecture series on algorithms and error correction.
A thorough review of the most commercially plausible application: mapping electronic structure onto qubits, and the algorithms that then extract energies.
A thorough, notebook-based curriculum built on the fast Qulacs simulator, strongest on quantum chemistry and variational methods. Under-known outside Japan and better than most English equivalents.
Hybrid quantum-classical models built as Keras layers, for people who already know TensorFlow. The tutorials are the fastest route from ML experience into QML.
This unlocks
Topics that list Variational algorithms (VQE & QAOA) as a prerequisite.