الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 150 |  |  |
| | | from | 150 | | |
Abstract
Quantum machine learning is an exciting new field with the potential to revolu- tionize how information is learned and processed. It blends the power of quantum computing, harnessing the strange and wonderful laws of quantum mechanics, with the well-established techniques of machine learning. Quantum computing lever- ages phenomena like superposition and entanglement, allowing them to process information in fundamentally different ways. This translates to several advantages for the deployed machine learning tasks, such as significant speedups or enhanced accuracy compared to classical machine learning methods. On the other hand, quantum hardware technology is still maturing creating a large gap between some quantum algorithms resource requirement and the available hardware capability. To overcome both quantum and classical hardware limitations, this study depends on hybrid quantum-classical machine learning models with focus on classification problems. This new breed of hybrid models has proved their suitability to the currently available quantum hardware in terms of robustness to noise and their dependency on limited number of qubits.
The main contribution of this thesis is to propose the design of two different hybrid classifier models and evaluate their corresponding prediction accuracy values. The first model is a binary classifier. The proposed model has been evaluated using 4 different benchmark datasets. The correlation between the circuit descriptors and the prediction accuracy values is also studied. The second model is a Quan- tum Natural Language Processing (QNLP) model used to identify whether two sentences are referring to the same topic or not, also identified in the context as a topic-aware classifier. The experimental results showed computational advan- tage of the QNLP model in terms of enhanced classification accuracy over classical tensor-network based model, the QNLP model also showed resilience against quan- tum hardware noise.