Entangled Quantum Neural Network

Quantum entanglement (QE) is the phenomenon that when several particles interact, the properties of each particle will be integrated into the properties of the overall system, and the properties of each particle cannot be described independently from others. QE can be proved by violating Bell Inequality, that is, it can describe strong statistical correlation (i.e., quantum correlation). By introducing QE into machine learning, the adjusted framework would have advantages such as faster execution time of the learning algorithms and stronger capacity. Therefore, here we introduce our novel framework called Entangled Quantum Neural Network, EQNN. Using quantum entanglement to development on neuron networks can be described in three different ways: By replacing the hidden layer nodes of the Multi-Layer Perception (MLP) with a measurement process of entangled states (QECA, QCCA); by replacing the output layer of MLP with a quantum measurement operation (ECA); or by reconstructing the neurons in NN using regularizer to constrain state vectors to entangled states (QNN). With extensive experiments on the three most frequently used machine learning datasets from UCI, Abalone, Wine Quality (Red), and Wine Quality (White), we demonstrate that all QCCA, QECA, ECA, and QNN outperform the baseline algorithms. Under the same parameter settings, which are: learning rate is 0.001, mini-batch is 1, training epochs is 500 and initial weight is 0.01, the performance among themselves in descending order are QNN > ECA > QECA > QCCA.