Implementation of dimensional reduction and classifier based on quantum programming

Quantum computing technology, with a different architecture from traditional computers, is rapidly advancing. In recent years, there have been remarkable efforts in developing programs suitable for quantum computers, computing environments, and preparing necessary algorithms. The limited accessibility to existing quantum computers leads researchers to extensively utilize quantum simulators in their analyses. These simulators also showcase performance variations. The choice of dimension reduction method, dataset size, and analysis environment are evidently significant factors influencing the execution process of studies. Ongoing developments in quantum programming manage to compete with classical methods despite existing limitations. This thesis extensively delves into the amalgamation of quantum computation and artificial intelligence, specifically focusing on quantum machine learning and algorithms. The study primarily encompasses pivotal dimension reduction and classification processes within the realm of artificial intelligence. Unlike classical classification algorithms, the input data for executing quantum machine learning algorithms needs to conform to computational environments. Due to the limited qubit sizes in quantum computing environments, dimension reduction methods are widely employed in this domain. This thesis conducts extensive applications concerning classical dimension reduction methods such as PCA and LDA. Medical datasets related to cardiology, diabetes, and obesity medicine were utilized for the quantum machine learning algorithms. Comprehensive studies were conducted under various scenarios on these datasets. Classical dimension reduction methods were applied to the datasets to create feature maps, generating datasets suitable for qubit usage. The Quantum Support Vector Machine (QSVM), a quantum machine learning algorithm, was implemented on these new datasets. To evaluate the performance of the QSVM algorithm, analyses were conducted with commonly used classical classification algorithms, including the classical Support Vector Machine, from the literature. The results obtained from the analyses were scrutinized in terms of performance and time. Various metrics pertaining to the performance of the QSVM algorithm were presented based on the chosen dimension reduction method. The conducted studies showcased promising results for classical dimension reduction methods and the QSVM approach on medical datasets. This thesis significantly contributes to the field of quantum machine learning through comprehensive applications.