The state of quantum learning: A comparative review towards classical machine learning

Abstract

This article examines the effectiveness of quantum machine learning and compares it with classical learning. Classical learning works on bits 0 and 1, but quantum learning is a superposition state of qubits |0 and |1 at the same time. Quantum machine learning speeds up the transductive learning method due to the adiabatic computing algorithm, where classical ones fail. In this regard, its speed is exponential and is attributed to Grover's search, Hamming distance, and quantum associative memory. Quantum machine learning offers parallelism in pattern recognition of data instances. It is absent in classical learning. The difference between quantum neural networks and classical ones is that interference is absent at the latter's perceptrons. The interference nature allows quantum neural networks to generalize patterns in databases effectively.