Harnessing The Power Of Quantum Algorithms For Solving Complex Data Science

Abstract

Quantum computing has become the possible paradigm to address exponentially growing complexity of data-driven algorithms which current classical algorithms are becoming unequipped to handle effectively (Montanaro, 2016). Although the conventional machine learning and deep learning algorithms have been very successful in various fields, they tend to fail to scale when they encounter large dimensions of data and combinatorial cases of optimization models (Biamonte et al., 2017). This paper will examine how quantum algorithms may be combined with data science problems, to fill the divide between potential and actual applications. Based on quantum machine learning, including the Variational Quantum Eigensolver (Peruzzo et al., 2014) and Quantum Support Vector Machines (Rebentrost et al., 2014), this work provides an introduction on how close quantum-classical models are able to exhibit better-than-classical performance in a series of synthetic and benchmark datasets. The achievement with our results can be summarized as computationally scalable and solution quality enhancement especially on clustering and optimization problems. This contribution to the literature helps to address the recent paradigm change to consider quantum computing as an effective technology capable of enhancing data science by introducing a pipelining that was reproducible and a focus on algorithmic fairness (Ciliberto et al., 2018). The results open up the road to the future research on vicarious scalable QML solutions to real-world problems and large-scale problems.

Keywords: Quntum computing et lunch quantum algorithms & data science, machine learning &

optimization & clustering & hybrid models.