@ericyocamShipped: Quantum Image Classification — Cirq vs. PennyLane
An open-source quantum machine learning proof of concept exploring handwritten digit classification using quantum circuits and variational classifiers.
The project uses the MNIST dataset to demonstrate two quantum computing development approaches: Google's Cirq framework and PennyLane.
The Cirq notebook constructs quantum circuits and illustrates their operations while also building and training a classical TensorFlow/Keras image classifier.
The PennyLane notebook implements a variational quantum classifier using parameterized quantum circuits, RX rotation gates, StronglyEntanglingLayers, and Pauli-Z expectation-value measurements.
The repository includes executable Jupyter notebooks, circuit demonstrations, and classification examples to help researchers and developers explore how quantum circuits can be integrated into machine learning workflows.
Designed for AI researchers, quantum computing developers, students, and educators investigating quantum machine learning and hybrid computational approaches.
The experiments are educational demonstrations and do not establish a quantum advantage over conventional machine learning classifiers.
2 comments
Package the two notebooks as a ready-to-teach two-week module, with slides, a homework set and an answer key, then send it to professors at the other Cal State campuses who want to add quantum to a class but have no materials. Your note that this doesn't show a quantum advantage helps here, because an instructor can teach it honestly without overselling it to students.
Great suggestion! I actually have a book on cybersecurity, quantum computing, and AI under contract with a publisher, co-authored with three other researchers. Turning these notebooks into a two-week teaching module would complement the book nicely. Providing ready-to-use lectures, hands-on exercises, and assessments could make quantum machine learning more accessible to faculty and students. And I agree that teaching the concepts without overstating quantum advantage is important. Appreciate the suggestion!