Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
1,295
result(s) for
"Nonlinear systems Data processing."
Sort by:
Applications of artificial neural networks for nonlinear data
\"This book is a collection of research on the contemporary nature of artificial neural networks and their specific implementations within data analysis\"-- Provided by publisher.
Book
Nonlinear optical encoding enabled by recurrent linear scattering
Optical information processing and computing can potentially offer enhanced performance, scalability and energy efficiency. However, achieving nonlinearity—a critical component of computation—remains challenging in the optical domain. Here we introduce a design that leverages a multiple-scattering cavity to passively induce optical nonlinear random mapping with a continuous-wave laser at a low power. Each scattering event effectively mixes information from different areas of a spatial light modulator, resulting in a highly nonlinear mapping between the input data and output pattern. We demonstrate that our design retains vital information even when the readout dimensionality is reduced, thereby enabling optical data compression. This capability allows our optical platforms to offer efficient optical information processing solutions across applications. We demonstrate our design’s efficacy across tasks, including classification, image reconstruction, keypoint detection and object detection, all of which are achieved through optical data compression combined with a digital decoder. In particular, high performance at extreme compression ratios is observed in real-time pedestrian detection. Our findings open pathways for novel algorithms and unconventional architectural designs for optical computing.
An optical accelerator is designed to leverage a multiple-scattering cavity to passively induce optical nonlinear random mapping with a continuous-wave laser at a constant low power (~21 mW), providing a new avenue for optical computing.
Journal Article
Chemical reservoir computation in a self-organizing reaction network
Chemical reaction networks, such as those found in metabolism and signalling pathways, enable cells to process information from their environment
1
,
2
. Current approaches to molecular information processing and computation typically pursue digital computation models and require extensive molecular-level engineering
3
. Despite considerable advances, these approaches have not reached the level of information processing capabilities seen in living systems. Here we report on the discovery and implementation of a chemical reservoir computer based on the formose reaction
4
. We demonstrate how this complex, self-organizing chemical reaction network can perform several nonlinear classification tasks in parallel, predict the dynamics of other complex systems and achieve time-series forecasting. This in chemico information processing system provides proof of principle for the emergent computational capabilities of complex chemical reaction networks, paving the way for a new class of biomimetic information processing systems.
A chemical reservoir computer based on the formose reaction has been discovered that can perform several nonlinear classification tasks in parallel, predict the dynamics of other complex systems and achieve time-series forecasting.
Journal Article
Deep learning prediction of noise-driven nonlinear instabilities in fibre optics
Machine learning is bringing revolutionary approaches into many fields of physics. Among those, photonics enables fast and scalable information processing. Photonics platforms further possess rich nonlinear dynamics that drive fundamental interest but also prove powerful for applications in computation, imaging, frequency conversion, source development and advanced signal processing. However, incoherent processes of nonlinear optics are hardly exploited in practice as the control of noise-driven dynamics remains challenging. Here, we exploit deep learning strategies and demonstrate that coherent optical seeding can effectively shape incoherent spectral broadening. We focus on the intricate interplay between weak coherent pulses and broadband noise, competing during nonlinear fibre propagation within an amplification process known as modulation instability. We demonstrate artificial neural networks’ capability to efficiently predict these complex incoherent dynamics, both numerically and experimentally. Our results show that input seed properties can be inferred from the incoherent output signal. Furthermore, our approach enables reliable prediction of output spectral fluctuations, paving the way to tailoring complex photonic signals with specific correlation features.
Neural networks unlock the ability to predict and manipulate noisy light propagation in nonlinear fibres, advancing control strategies for photonic technologies and paving the way for tailored shaping of incoherent optical signals.
Journal Article
Quantized neural adaptive finite-time preassigned performance control for interconnected nonlinear systems
In this article, the issue of neural adaptive decentralized finite-time prescribed performance (FTPP) control is investigated for interconnected nonlinear time-delay systems. First, to bypass the potential singularity difficulties, the hyperbolic tangent function and the radial basis function neural networks are integrated to handle the unknown nonlinear items. Then, an adaptive FTPP control strategy is developed, where an improved fractional-order filter is applied to tackle the tremendous “amount of calculation” and eliminate the filter error simultaneously. Furthermore, by considering the impact of bandwidth limitation, an adaptive self-triggered control law is designed, in which the next trigger instant is determined through the current information. Ultimately, it can be demonstrated that the proposed control scheme not only guarantees that all states of the closed-loop system are semi-globally uniformly ultimately bounded, but also that the system output is confined to a small area in finite time. Two simulation examples are carried out to verify the effectiveness and superiority of the proposed method.
Journal Article
Large optical nonlinearity of indium tin oxide in its epsilon-near-zero region
Nonlinear optical phenomena are crucial for a broad range of applications, such as microscopy, all-optical data processing, and quantum information. However, materials usually exhibit a weak optical nonlinearity even under intense coherent illumination. We report that indium tin oxide can acquire an ultrafast and large intensity-dependent refractive index in the region of the spectrum where the real part of its permittivity vanishes. We observe a change in the real part of the refractive index of 0.72 ± 0.025, corresponding to 170% of the linear refractive index. This change in refractive index is reversible with a recovery time of about 360 femtoseconds. Our results offer the possibility of designing material structures with large ultrafast nonlinearity for applications in nanophotonics.
Journal Article
Deep learning with coherent VCSEL neural networks
Deep neural networks (DNNs) are reshaping the field of information processing. With the exponential growth of these DNNs challenging existing computing hardware, optical neural networks (ONNs) have recently emerged to process DNN tasks with high clock rates, parallelism and low-loss data transmission. However, existing challenges for ONNs are high energy consumption due to their low electro-optic conversion efficiency, low compute density due to large device footprints and channel crosstalk, and long latency due to the lack of inline nonlinearity. Here we experimentally demonstrate a spatial-temporal-multiplexed ONN system that simultaneously overcomes all these challenges. We exploit neuron encoding with volume-manufactured micrometre-scale vertical-cavity surface-emitting laser (VCSEL) arrays that exhibit efficient electro-optic conversion (<5 attojoules per symbol with a π-phase-shift voltage of Vπ = 4 mV) and compact footprint (<0.01 mm2 per device). Homodyne photoelectric multiplication allows matrix operations at the quantum-noise limit and detection-based optical nonlinearity with instantaneous response. With three-dimensional neural connectivity, our system can reach an energy efficiency of 7 femtojoules per operation (OP) with a compute density of 6 teraOP mm−2 s−1, representing 100-fold and 20-fold improvements, respectively, over state-of-the-art digital processors. Near-term development could improve these metrics by two more orders of magnitude. Our optoelectronic processor opens new avenues to accelerate machine learning tasks from data centres to decentralized devices.Energy consumption and compute density are challenges for computing systems. Here researchers show an optical computing architecture using micrometre-scale VCSEL transmitter arrays enabling 7 fJ energy per operation and a potential compute density of 6 tera-operations mm−2 s−1.
Journal Article
Adaptive fuzzy fault-tolerant control using Nussbaum-type function with state-dependent actuator failures
This paper presents an adaptive fuzzy fault-tolerant tracking control for a class of unknown multi-variable nonlinear systems, with external disturbances, unknown control sign, and actuator faults. By employing fuzzy logic systems, the unknown nonlinear dynamics and the state-dependent actuator faults are approximated, and by utilizing a Nussbaum-type function, the issue of unknown control sign is solved. The proposed control scheme is based on two forms, an adaptive fuzzy controller along with a robust controller that is equipped with a Nussbaum-type gain function, which guarantees stability with the boundedness of all signals involved in the closed-loop system. To prove the accuracy, and the effectiveness of the proposed control scheme, a simulation example on two-inverted pendulums system is carried out.
Journal Article
Nonlinear optics in the fractional quantum Hall regime
Engineering strong interactions between optical photons is a challenge for quantum science. Polaritonics, which is based on the strong coupling of photons to atomic or electronic excitations in an optical resonator, has emerged as a promising approach to address this challenge, paving the way for applications such as photonic gates for quantum information processing
1
and photonic quantum materials for the investigation of strongly correlated driven–dissipative systems
2
,
3
. Recent experiments have demonstrated the onset of quantum correlations in exciton-polariton systems
4
,
5
, showing that strong polariton blockade
6
—the prevention of resonant injection of additional polaritons in a well delimited region by the presence of a single polariton—could be achieved if interactions were an order of magnitude stronger. Here we report time-resolved four-wave-mixing experiments on a two-dimensional electron system embedded in an optical cavity
7
, demonstrating that polariton–polariton interactions are strongly enhanced when the electrons are initially in the fractional quantum Hall regime. Our experiments indicate that, in addition to strong correlations in the electronic ground state, exciton–electron interactions leading to the formation of polaron-polaritons
8
–
11
have a key role in enhancing the nonlinear optical response of the system. Our findings could facilitate the realization of strongly interacting photonic systems, and suggest that nonlinear optical measurements could provide information about fractional quantum Hall states that is not accessible through their linear optical response.
Nonlinear optical measurements in a two-dimensional electron system embedded in an optical cavity show enhanced polariton–polariton interactions in the fractional quantum Hall regime.
Journal Article
Biological neurons act as generalization filters in reservoir computing
Reservoir computing is a machine learning paradigm that transforms the transient dynamics of high-dimensional nonlinear systems for processing time-series data. Although the paradigm was initially proposed to model information processing in the mammalian cortex, it remains unclear how the nonrandom network architecture, such as the modular architecture, in the cortex integrates with the biophysics of living neurons to characterize the function of biological neuronal networks (BNNs). Here, we used optogenetics and calcium imaging to record the multicellular responses of cultured BNNs and employed the reservoir computing framework to decode their computational capabilities. Micropatterned substrates were used to embed the modular architecture in the BNNs. We first show that the dynamics of modular BNNs in response to static inputs can be classified with a linear decoder and that the modularity of the BNNs positively correlates with the classification accuracy. We then used a timer task to verify that BNNs possess a short-term memory of several 100 ms and finally show that this property can be exploited for spoken digit classification. Interestingly, BNN-based reservoirs allow categorical learning, wherein a network trained on one dataset can be used to classify separate datasets of the same category. Such classification was not possible when the inputs were directly decoded by a linear decoder, suggesting that BNNs act as a generalization filter to improve reservoir computing performance. Our findings pave the way toward a mechanistic understanding of information representation within BNNs and build future expectations toward the realization of physical reservoir computing systems based on BNNs.
Journal Article