research / quantum-machine-learning-thesis · B.Sc. thesis (CSE499), North South University · 2022
Machine Learning In The Realm Of Quantum: The State-Of-The-Art, Challenges, Future Vision and Applications Of It
Md. Abu Ammar, Sadia Afrin Tamanna · supervised by Dr. Mahdy Rahman Chowdhury
[quantum-ml][quanvolution][cvqnn][pennylane][mnist]
Abstract
A comprehensive review of the state of the art in quantum machine learning, paired with hands-on classification experiments: two first-generation hybrid quantum-classical models — a quanvolutional neural network on a gate-based simulator and a continuous-variable quantum neural network on a photonic simulator — trained on MNIST and compared head-to-head against classical baselines on accuracy and convergence.
§1In plain words
Can today's quantum computers learn to read handwritten digits? This thesis builds two very different quantum learners to find out. The first slides a tiny 4-qubit circuit across each image the way a convolutional filter would — a quanvolution. The second encodes images into beams of light (squeezers, beamsplitters, Kerr gates) following Xanadu's continuous-variable recipe. Both are trained the same way a normal neural network is, with gradients flowing through the quantum circuit via the parameter-shift rule — the same mathematics running live in this site's hero and in lesson 6 of /learn.
§2Method
PennyLane with the Keras plugin, 10-class MNIST. Model 1 (quanvolutional, after Henderson et al. 2019): 2×2 patches angle-encoded via RY rotations into 4 qubits, a random variational layer, Pauli-Z expectations giving 4 feature channels, then a softmax head. Model 2 (CV-QNN, after Killoran et al.): dense classical layers compress each image to 14 parameters driving a 2-qumode photonic circuit — squeezing, interferometers, displacement, Kerr nonlinearity — with 4 quantum layers (56 quantum parameters), trained on a 700-sample subset with cutoff dimension 4.
§3Results
The classical baselines won — and the margins are the finding. The quanvolutional model reached 92% test accuracy against a 96% classical CNN, and converged to its optimum loss faster than the classical model. The CV-QNN reached 72% against an 88% classical baseline. First-generation QML models trained on classical data did not beat optimized classical networks, but the gate-based approach came close — the honest state of the NISQ era, measured directly.
§4Looking back
The thesis taught me that the interesting question isn't "is quantum faster?" but "where does the encoding bottleneck bite?" Getting classical pixels into a quantum state dominated every design decision. That lesson shaped this site: the interactive curriculum at /learn exists because the encoding intuition took me months and a 68-page thesis to build, and a visitor can now get it in six scroll-stops.
@thesis{ammar2022qml,
author = {Ammar, Md. Abu and Tamanna, Sadia Afrin},
title = {Machine Learning In The Realm Of Quantum: The
State-Of-The-Art, Challenges, Future Vision and
Applications Of It},
type = {Bachelor's thesis},
institution = {North South University},
address = {Dhaka, Bangladesh},
year = {2022},
note = {Supervised by Dr. Mahdy Rahman Chowdhury}
}