Machine learning now plays a pivotal role in our society, providing solutions to problems that were previously thought intractable. The meteoric rise of this technology can no doubt be attributed to the information age that we now live in. As data is continually amassed, more efficient and scalable methods are required to yield functional models and accurate inferences. Simultaneously we have also seen quantum technology come to the forefront of research and next generation systems. These technologies promise secure information transfer, efficient computation and high precision sensing, at levels unattainable by their classical counterparts. Although these technologies are powerful, they are necessarily more complicated and difficult to control. The combination of these two advances yields an opportunity for study, namely leveraging the power of machine learning to control and optimise quantum (and more generally complex) systems. The work presented in thesis explores these avenues of investigation and demonstrates the potential success of machine learning methods in the domain of quantum and complex systems. One of the most crucial potential quantum technologies is the quantum memory. If we are to one day harness the true power of quantum key distribution for secure transimission of information, and more general quantum computating tasks, it will almost certainly involve the use of quantum memorys. We start by presenting the operation of the cold atom workhorse: the magneto-optical trap (MOT). To use a cold atomic ensemble as a quantum memory we are required to prepare the atoms using a specialised cooling sequence. During this we observe a stable, coherent optical emission exiting each end of the elongated ensemble. We characterise this behaviour and compare it to similar observations in previous work. Following this, we use the ensemble to implement a backward Raman memory. Using this scheme we are able to demonstrate an increased efficiency over that of previous forward recall implementations. While we are limited by the optical depth of the system, we observe an efficiency more than double that of previous implementations. The MOT provides an easily accessible test bed for the optimisation via some machine learning technique. As we require an efficient search method, we implement a new type of algorithm based on deep learning. We design this technique such that the artificial neural networks are placed in control of the online optimisation, rather than simply being used as surrogate models. We experimentally optimise the optical depth of the MOT using this method, by parametrising the time varying compression sequence. We identify a new and unintuitive method for cooling the atomic ensemble which surpasses current methods. Following this initial implementation we make substantial improvements to the deep learning approach. This extends the approach to be applicable to a far wider range of complex problems, which may contain extensive local minima and structure. We benchmark this algorithm against many of the conventional optimisation techniques and demonstrate superior capability to optimise problems with high dimensionality. Finally we apply this technique to a series of preliminary problems, namely the tuning of a single electron transistor and second-order correlations from a quantum dot source.
Machine learning now plays a pivotal role in our society, providing solutions to problems that were previously thought intractable. The meteoric rise of this technology can no doubt be attributed to the information age that we now live in. As data is continually amassed, more efficient and scalable methods are required to yield functional models and accurate inferences. Simultaneously we have also seen quantum technology come to the forefront of research and next generation systems. These technologies promise secure information transfer, efficient computation and high precision sensing, at levels unattainable by their classical counterparts. Although these technologies are powerful, they are necessarily more complicated and difficult to control. The combination of these two advances yields an opportunity for study, namely leveraging the power of machine learning to control and optimise quantum (and more generally complex) systems. The work presented in thesis explores these avenues of investigation and demonstrates the potential success of machine learning methods in the domain of quantum and complex systems. One of the most crucial potential quantum technologies is the quantum memory. If we are to one day harness the true power of quantum key distribution for secure transimission of information, and more general quantum computating tasks, it will almost certainly involve the use of quantum memorys. We start by presenting the operation of the cold atom workhorse: the magneto-optical trap (MOT). To use a cold atomic ensemble as a quantum memory we are required to prepare the atoms using a specialised cooling sequence. During this we observe a stable, coherent optical emission exiting each end of the elongated ensemble. We characterise this behaviour and compare it to similar observations in previous work. Following this, we use the ensemble to implement a backward Raman memory. Using this scheme we are able to demonstrate an increased efficiency over that of previous forward recall implementations. While we are limited by the optical depth of the system, we observe an efficiency more than double that of previous implementations. The MOT provides an easily accessible test bed for the optimisation via some machine learning technique. As we require an efficient search method, we implement a new type of algorithm based on deep learning. We design this technique such that the artificial neural networks are placed in control of the online optimisation, rather than simply being used as surrogate models. We experimentally optimise the optical depth of the MOT using this method, by parametrising the time varying compression sequence. We identify a new and unintuitive method for cooling the atomic ensemble which surpasses current methods. Following this initial implementation we make substantial improvements to the deep learning approach. This extends the approach to be applicable to a far wider range of complex problems, which may contain extensive local minima and structure. We benchmark this algorithm against many of the conventional optimisation techniques and demonstrate superior capability to optimise problems with high dimensionality. Finally we apply this technique to a series of preliminary problems, namely the tuning of a single electron transistor and second-order correlations from a quantum dot source.
Machine learning based on artificial neural networks has emerged as an efficient means to develop empirical models of complex systems. Cold atomic ensembles have become commonplace in laboratories around the world, however, many-body interactions give rise to complex dynamics that preclude precise analytic optimisation of the cooling and trapping process. Here, we implement a deep artificial neural network to optimise the magneto-optic cooling and trapping of neutral atomic ensembles. The solution identified by machine learning is radically different to the smoothly varying adiabatic solutions currently used. Despite this, the solutions outperform best known solutions producing higher optical densities.
Abstract Machine learning based on artificial neural networks has emerged as an efficient means to develop empirical models of complex systems. Cold atomic ensembles have become commonplace in laboratories around the world, however, many-body interactions give rise to complex dynamics that preclude precise analytic optimisation of the cooling and trapping process. Here, we implement a deep artificial neural network to optimise the magneto-optic cooling and trapping of neutral atomic ensembles. The solution identified by machine learning is radically different to the smoothly varying adiabatic solutions currently used. Despite this, the solutions outperform best known solutions producing higher optical densities.
Quantum optical systems are poised to become integral components of technologies of the future. While there is growing commercial interest in these systems---for applications in information processing, secure communication and precision metrology---there remain significant technical challenges to overcome before widespread adoption is possible. In this thesis we consider the general problem of optimising quantum optical systems, with a focus on sensing and information processing applications. We investigate four different classes of system with varying degrees of generality and complexity, and demonstrate four corresponding optimisation techniques. At the most specific end of the spectrum---where behaviour is best understood---we consider the problem of interferometric sensitivity enhancement, specifically in the context of long-baseline gravitational wave detectors. We investigate the use of an auxiliary optomechanical system to generate squeezed light exhibiting frequency-dependent quadrature rotation. Such rotation is necessary to evade the effect of quantum back action and achieve broadband sensitivity beyond the standard quantum limit. We find that a cavity optomechanical system is generally unsuitable for this purpose, since the quadrature rotation occurs in the opposite direction to that required for broadband sensitivity improvement. Next we introduce a general technique to engineer arbitrary optical spring potentials in cavity optomechanical systems. This technique has the potential to optimise many types of sensors relying on the optical spring effect. As an example, we show that this technique could yield an enhancement in sensitivity by a factor of 5 when applied to a certain gravitational sensor based on a levitated cavity mirror. We then consider a particular nanowire-based optomechanical system with potential applications in force sensing. We demonstrate a variety of ways to improve its sensitivity to transient forces. We first apply a non-stationary feedback cooling protocol to the system, and achieve an improvement in peak signal-to-noise ratio by a factor of 3, corresponding to a force resolution of 0.2fN. We then implement two non-stationary estimation schemes, which involve post-processing data taken in the absence of physical feedback cooling, to achieve a comparable enhancement in performance without the need for additional experimental complexity. Finally, to address the most complex of systems, we present a general-purpose machine learning algorithm capable of automatically modelling and optimising arbitrary physical systems without human input. To demonstrate the potential of the algorithm we apply it to a magneto-optical trap used for a quantum memory, and achieve an improvement in optical depth from 138 to 448. The four techniques presented differ significantly in their style and the types of systems to which they are applicable. Successfully harnessing the full range of such optimisation procedures will be vital in unlocking the potential of quantum optical systems in the technologies of the future
The development of quantum technologies has been an important topic of research and technological development over the last few years, with numerous projects aiming to exploit cold atoms and Bose-Einstein condensates for sensing applications and probes of fundamental physics. To further develop the emerging applications of new quantum mechanical phenomena, the UK established the a national quantum tech program in 2014. The University of Nottingham, alongside university partners across the country, is currently involved in advancing quantum-based technologies for sensors and metrology applications. A key part of many experiments in this area is the Magneto-Optical Trap, or MOT, which plays a pivotal role in the initial stages of the atom cooling required to reach the nK temperatures necessary to form a condensate. Traditionally, MOTs are typically comprised of multiple pieces of specialised equipment including an ultra-high vacuum chamber, several lasers and magnetic field generating devices, all of which occupy 10's cm in linear dimensions on a laboratory bench and require mains power. As in previous quantum-enabled technologies, including MRI scanners, simulation and optimisation may prove crucial in reducing the size, weight and power requirements of such systems, and for improving their performance. Consequently, it is natural to apply such optimisation methods to emerging quantum technologies, in order to meet the restrictions on experimental size, weight and power necessary the practical cold atom-based sensing systems. As it forms a crucial part of many cold atom quantum technologies, the MOT is a natural initial target for simulation and subsequent optimisation. To develop system design capabilities, this thesis firstly presents three optimisation methods for permanent magnets. One of these builds on previous work to optimise ferromagnetic films for Ioffe-Pritchard arrays (chapter \ref{Magnetic Plate Optimisation}), the others are novel optimisation methods, drawing on more general approaches and on shimming methods frequently used in MRI. Permanent magnets were chosen for use in MOTs in order to reduce the required power and volume of current experimental equipment. While each of the three optimisation methods have different capabilities, two were particularly powerful for designing high field fidelity MOTs. One of these methods was developed specifically to design magnet systems for use with the 3D printed vacuum chamber of the Innovate UK-funded OptaMOT project, illustrating its potential to develop experimental systems. The magnetic field was generated to high accuracy by using the new optimisation algorithm, presented in chapter \ref{Spherical Harmonic Optimisation}, by using mass produced magnets, thereby offering improved practicality over designs that require, expensive, custom-made magnets. With the aim of improving the holistic design capabilities required to advance MOTs, simulation software for the dynamics of atoms in MOTs was also developed. The simulations were capable of modelling the dynamics of atom clouds in both a typical 6 beam MOT and an optical molasses arrangement. It was also used to elucidate the action of a significantly more complicated beam arrangement exploiting more advanced physical phenomena. Remarkably, the arrangement, developed by Hub partners at The University of Southampton, experimentally shows a high atom number and the potential presence of a restoring force without an applied magnetic field. The laser beam geometry also exhibits an unusual dependence of the final atom number on laser beam detuning and power. The simulations provided insights into the potential origins of this dependence (suggesting a combination of beam interference and an additional detuning from the laser fields), whilst confirming that the laser beam configuration alone creates a damping force, as with an optical molasses arrangement. However, the origins and presence of a restoring force still, unfortunately, remain unknown.
Quantum reservoir computing has emerged as a promising paradigm within the field of quantum machine learning, harnessing the inherent properties of quantum systems to optimise and enhance information processing capabilities. Here, we explore the potential of quantum-inspired machine learning methodologies by leveraging the complex dynamics of quantum reservoirs to address computationally challenging tasks with enhanced efficiency and accuracy. To this end, we employ an open quantum system model comprising two-level atomic ensembles coupled to Lorentzian photonic cavities to construct a quantum physical reservoir computer layer for a recurrent neural network. We evaluate the effectiveness of this approach by applying it to a standard machine learning image-recognition problem and benchmarking its performance against a conventional neural network of similar architecture, but lacking the quantum physical reservoir computer layer. Remarkably, as the dataset size increases, the quantum physical reservoir computer outperforms the conventional neural network, requiring fewer training epochs and a smaller dataset to achieve comparable accuracy. Furthermore, we employ the quantum physical reservoir computing approach to model the dynamics of open quantum systems, focusing on atomic system ensembles interacting with a structured photonic reservoir associated with a photonic band-gap material. Our results reveal that the quantum reservoir computer provides equally powerful representations for quantum dynamical problems, maintaining effectiveness even under constraints of limited training data.
[eng] Quantum photonic technologies have achieved considerable success across a range of disciplines, including quantum computing, quantum communications, and quantum metrology. Conversely, in the current era of information, the ability to discern patterns within voluminous data sets is highly sought after, with Machine Learning (ML) having emerged as a powerful solution to address this challenge. The processing speed of photonic hardware, in conjunction with the potential for computation enhancement resulting from quantum properties, positions quantum photonics as a promising platform for the implementation of ML algorithms. In this thesis, we focus on reservoir computing (RC) as an ML framework that is suitable for time series processing. It provides fast and energy-efficient training in comparison to more conventional methods and is also appropriate for physical implementations. In this thesis, we explore the possibilities brought by quantum photonic setups in the field of quantum RC. Specifically, we propose and study different platform designs that are able to perform different temporal benchmark tasks by providing tunable memory and nonlinearity. Furthermore, we leverage the capabilities of state-of-the-art quantum photonic hardware and identify novel strategies to address key challenges in RC in quantum substrates, as well as expand the potential of quantum machine learning. Firstly, an approach is provided to overcome some of the limitations of quantum projective measurements, thereby enabling information processing in real time. This is achieved by generating a physical ensemble of quantum reservoirs that recirculate through an optical fiber inside a feedback loop. We investigate the detrimental effects of sampling noise due to finite ensembles and provide a strategy to mitigate them. Secondly, the consequences of increasing quantum squeezing in the dynamics of such setups were studied, and it was found that it monotonically improves the noise robustness of the platforms in realistic scenarios. Finally, we propose a novel design in which photonic cluster states, a specific kind of highly entangled quantum network, are used as resources to implement quantum RC. In this configuration, external input information is encoded into the quantum network via quantum teleportation by engineering the measurement basis of the protocol. Furthermore, the platform has the capacity to implement global quantum gates through local operations (single-mode measurements), thus establishing the foundation for distributed quantum ML.