Pennylane and Qiskit for quantum machine learning

Question

I'm interested in quantum computing, specifically in “quantum machine learning” (QML). I'm going to start my masters program in computer science and have previous experience in classical machine learning. I'd like to learn quantum mechanics generally but focus on quantum algorithms.

I have read about Qiskit and Pennylane, but I'm unsure where to look next. Could you recommend books, courses, and papers to gain background into QML? My goal is to use Qiskit and Pennylane to develop QML algorithms in NISQ systems for data science and big data.

Another question - can we use quantum computing to develop Bayesian deep learning models (Gaussian processes, bayesian quantum circuits) for data science and big data? Thanks in advance.

Answer 1

Have a look at these for quantum machine learning:

• Supervised learning with quantum computers by Schuld and Petruccione (2018)
• An introduction to quantum machine learning by the same authors of the textbook above
• Quantum machine learning published in Nature 2017 by some experts in the field: Wittek, Rebentrost, Lloyd, et al
• Video presentations by Dr. Schuld and Dr. Wittek on quantum machine learning
• Quantum bayesian deep learning for Qiskit is discussed here

Answer 2

Since quantum machine learning with NISQ hardware is such a relatively new field, it is still very highly research driven, and a lot of the potential is still being determined.

To make these new research implementations more accessible, we've begun building implementations over at https://pennylane.ai/qml. Interesting ones include:

• Quantum Generative Adversarial Networks with Cirq + TensorFlow

• State preparation with Rigetti Forest + PyTorch

• Doubly stochastic gradient descent

• Variational Quantum Linear Solver

• QAOA for MaxCut

• Quantum variational classifier

These implementations are executable, and can be downloaded as Python scripts or Jupyter notebooks. Our goal is to keep regularly adding new implementations and demos (the repository behind the website is also open-source at https://github.com/XanaduAI/qml, so anyone can contribute demos).

Answer 3

You can also run QML/QAI on a real Quantum Computer using Qiskit.

Here are two sample Jupyter Notebooks for qSVM and qGAN. https://quantum-computing.ibm.com/jupyter/tutorial/advanced/aqua/artificial_intelligence/index.ipynb