Current Applicability of Quantum Machine Learning to Data Analytics

In recent years, quantum computing and its application to machine learning have evolved to the point where the data analytics practitioner must ask whether the technology is ready to aid large scale data processing tasks. This research describes the state of the art along with the limitations of error prone current generation quantum processors in the so called NISQ (Noisy Intermediate Stage Quantum) era. An introduction to the cloud based IBM Quantum infrastructure and Qiskit software application development kit is given. The design and implementation in software of quantum classifiers based on the distance measurement between quantum states using the SWAP Test circuit is examined. A method of using quantum feature maps and quantum kernel estimation for use with SVM classifiers on conventional computers is described. This research illustrates the feasibility of application of current quantum hardware technology to multi-dimensional datasets with a reasonable runtime performance. Future directions for research into suitable datasets and novel parameterized quantum feature map circuits is advised to empirically demonstrate advantage for fast, accurate classifications over conventional ML methods. This process will be aided by the ongoing improvement in quantum hardware capabilities and larger scale access to quantum computing services