Publication

Abstract : Entanglement detection and characterization in multiqubit systems is a difficult and challenging problem in quantum information processing, and existing analytical methods are limited to small Hilbert space dimensions. We design and successfully implement artificial neural networks (ANNs) to detect and classify entanglement for three-qubit systems using limited state features. The overall design principle is a feed-forward neural network, with the output layer consisting of a single neuron for the detection of genuine multipartite entanglement (GME) and six neurons for the classification problem corresponding to six entanglement classes under stochastic local operations and classical communication (SLOCC). The models are trained and validated on a simulated dataset of randomly generated states. We achieve high accuracy, around 98%, for detecting GME as well as for SLOCC classification. Remarkably, the ANN concludes that it requires only seven diagonal elements of the entire density matrix to detect GME as well as to classify the state into one of the six entanglement classes with an accuracy greater than 94% for both these tasks. This is an intriguing result, which does not lend itself easily to an intuitive explanation. Reducing the size of the feature set makes it easier to apply ANN models for entanglement classification, particularly in resource-constrained environments, without sacrificing accuracy. The performance of the ANN models is further evaluated by introducing white noise into the dataset, and the results indicate that the models are robust and are able to tolerate noise well.