Neural networks enjoy widespread success in both research and industry and, with the advent of quantum technology, it is a crucial challenge to design quantum neural networks for fully quantum learning tasks. Here we propose a truly quantum analogue of classical neurons, which form quantum feedforward neural networks capable of universal quantum computation. We describe the efficient training of these networks using the fidelity as a cost function, providing both classical and efficient quantum implementations. Our method allows for fast optimisation with reduced memory requirements: the number of qudits required scales with only the width, allowing deep-network optimisation. We benchmark our proposal for the quantum task of learning an unknown unitary and find remarkable generalisation behaviour and a striking robustness to noisy training data.

No abstract

Compared to light interferometers, the flux in cold-atom interferometers is low and the associated shot noise is large. Sensitivities beyond these limitations require the preparation of entangled atoms in different momentum modes. Here, we demonstrate a source of entangled atoms that is compatible with state-of-theart interferometers. Entanglement is transferred from the spin degree of freedom of a Bose-Einstein condensate to well-separated momentum modes, witnessed by a squeezing parameter of −3.1ð8Þ dB. Entanglement-enhanced atom interferometers promise unprecedented sensitivities for quantum gradiometers or gravitational wave detectors.

No abstract

Abstract. Problem. The operation of transport systems of a large city must ensure a continuous flow of passengers. The subway occupies a special place in the city's transport system. The constant increase in the loading of the subway not only worsens the conditions of passenger transportation, but also shortens the service life of the rolling stock and negatively affects the reliability of its operation. In the future, this may cause failures of the rolling stock, which means work failures. The main factor in the development of subway systems is passenger traffic, which determines the development of the subway network, the capacity of technical equipment, the number of vehicles, the size of depots and vehicle repair plants, and the time of repair with the least economic losses. Thus, one of the urgent tasks of the functioning of transport systems is the creation of an automated system of mass passenger service, in particular in the subway. Goal. The tasks of this study are the analysis of the subject area, the design of the information system according to the task, the selection of the necessary tools for this development, and the methods of software implementation. Analysis of passenger flows using an automated system will allow to solve many problems of design, operation, and planning of metro work, will allow to reduce the time for passenger service and optimize passenger flows. Methodology. Research methods are the methods of mass service systems. Results. The creation of a functional model will make it possible to develop a software product for the analysis and assessment of public transport passenger traffic. Originality.  Practical value. The software product created on the basis of the developed functional model will enable the user not only to analyze, but also to optimize the work of both existing and future metro stations.