Generalised Analog LSTMs Recurrent Modules for Neural Computing.
Ontology highlight
ABSTRACT:
SUBMITTER: Adam K
PROVIDER: S-EPMC8506007 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Analog machine learning hardware platforms promise to be faster and more energy efficient than their digital counterparts. Wave physics, as found in acoustics and optics, is a natural candidate for building analog processors for time-varying signals. Here, we identify a mapping between the dynamics of wave physics and the computation in recurrent neural networks. This mapping indicates that physical wave systems can be trained to learn complex features in temporal data, using standard training techniques for neural networks. As a demonstration, we show that an inverse-designed inhomogeneous medium can perform vowel classification on raw audio signals as their waveforms scatter and propagate through it, achieving performance comparable to a standard digital implementation of a recurrent neural network. These findings pave the way for a new class of analog machine learning platforms, capable of fast and efficient processing of information in its native domain.
| S-EPMC6924985 | biostudies-literature
Project description:Wave-based signal processing has witnessed a significant expansion of interest in a variety of science and engineering disciplines, as it provides new opportunities for achieving high-speed and low-power operations. Although flat optics desires integrable components to perform multiple missions, yet, the current wave-based computational metasurfaces can engineer only the spatial content of the input signal where the processed signal obeys the traditional version of Snell's law. In this paper, we propose a multi-functional metagrating to modulate both spatial and angular properties of the input signal whereby both symmetric and asymmetric optical transfer functions are realized using high-order space harmonics. The performance of the designed compound metallic grating is validated through several investigations where closed-form expressions are suggested to extract the phase and amplitude information of the diffractive modes. Several illustrative examples are demonstrated to show that the proposed metagrating allows for simultaneous parallel analog computing tasks such as first- and second-order spatial differentiation through a single multichannel structured surface. It is anticipated that the designed platform brings a new twist to the field of optical signal processing and opens up large perspectives for simple integrated image processing systems.
| S-EPMC9125804 | biostudies-literature
Project description:Acoustic computing devices, including switches, logic gates, differentiator and integrator, have attracted extensive attentions in both academic research and engineering. However, no scheme of acoustic computing device with more complex functionality has been proposed, such as ordinary differential equation (ODE) solver. Here, we propose an acoustic analog computing (AAC) system based on three cascaded metasurfaces to solve the nth-order ODEs. The metasurfaces are constructed with layered labyrinthine units featuring broad amplitude and phase modulation ranges. The simulated transmitted pressure of the AAC system agrees well with the theoretical solution of ODE, demonstrating the excellent functionality. Unlike the optical ODE solver based on differentiator or integrator, whose geometry becomes more complicated for solving higher order ODE, the proposed AAC system with fixed geometry can be designed for arbitrary nth-order ODE in principle. The proposal may find applications in various scenarios such as acoustic communication, analog computing and signal processing.
| S-EPMC6031674 | biostudies-literature
Project description:The electrical signals triggering the heart's contraction are governed by non-linear processes that can produce complex irregular activity, especially during or preceding the onset of cardiac arrhythmias. Forecasts of cardiac voltage time series in such conditions could allow new opportunities for intervention and control but would require efficient computation of highly accurate predictions. Although machine-learning (ML) approaches hold promise for delivering such results, non-linear time-series forecasting poses significant challenges. In this manuscript, we study the performance of two recurrent neural network (RNN) approaches along with echo state networks (ESNs) from the reservoir computing (RC) paradigm in predicting cardiac voltage data in terms of accuracy, efficiency, and robustness. We show that these ML time-series prediction methods can forecast synthetic and experimental cardiac action potentials for at least 15-20 beats with a high degree of accuracy, with ESNs typically two orders of magnitude faster than RNN approaches for the same network size.
| S-EPMC8502981 | biostudies-literature
Project description:Neurons communicate primarily with spikes, but most theories of neural computation are based on firing rates. Yet, many experimental observations suggest that the temporal coordination of spikes plays a role in sensory processing. Among potential spike-based codes, synchrony appears as a good candidate because neural firing and plasticity are sensitive to fine input correlations. However, it is unclear what role synchrony may play in neural computation, and what functional advantage it may provide. With a theoretical approach, I show that the computational interest of neural synchrony appears when neurons have heterogeneous properties. In this context, the relationship between stimuli and neural synchrony is captured by the concept of synchrony receptive field, the set of stimuli which induce synchronous responses in a group of neurons. In a heterogeneous neural population, it appears that synchrony patterns represent structure or sensory invariants in stimuli, which can then be detected by postsynaptic neurons. The required neural circuitry can spontaneously emerge with spike-timing-dependent plasticity. Using examples in different sensory modalities, I show that this allows simple neural circuits to extract relevant information from realistic sensory stimuli, for example to identify a fluctuating odor in the presence of distractors. This theory of synchrony-based computation shows that relative spike timing may indeed have computational relevance, and suggests new types of neural network models for sensory processing with appealing computational properties.
| S-EPMC3375225 | biostudies-literature
Project description:von Neumann architecture based computers isolate computation and storage (i.e. data is shuttled between computation blocks (processor) and memory blocks). The to-and-fro movement of data leads to a fundamental limitation of modern computers, known as the Memory wall. Logic in-Memory (LIM)/In-Memory Computing (IMC) approaches aim to address this bottleneck by directly computing inside memory units thereby eliminating energy-intensive and time-consuming data movement. Several recent works in literature, propose realization of logic function(s) directly using arrays of emerging resistive memory devices (example- memristors, RRAM/ReRAM, PCM, CBRAM, OxRAM, STT-MRAM etc.), rather than using conventional transistors for computing. The logic/embedded-side of digital systems (like processors, micro-controllers) can greatly benefit from such LIM realizations. However, the pure storage-side of digital systems (example SSDs, enterprise storage etc.) will not benefit much from such LIM approaches as when memory arrays are used for logic they lose their core functionality of storage. Thus, there is the need for an approach complementary to existing LIM techniques, that's more beneficial for the storage-side of digital systems; one that gives compute capability to memory arrays not at the cost of their existing stored states. Fundamentally, this would require memory nanodevice arrays that are capable of storing and computing simultaneously. In this paper, we propose a novel 'Simultaneous Logic in-Memory' (SLIM) methodology which is complementary to existing LIM approaches in literature. Through extensive experiments we demonstrate novel SLIM bitcells (1T-1R/2T-1R) comprising non-filamentary bilayer analog OxRAM devices with NMOS transistors. Proposed bitcells are capable of implementing both Memory and Logic operations simultaneously. Detailed programming scheme, array level implementation, and controller architecture are also proposed. Furthermore, to study the impact of proposed SLIM approach for real-world implementations, we performed analysis for two applications: (i) Sobel Edge Detection, and (ii) Binary Neural Network- Multi layer Perceptron (BNN-MLP). By performing all computations in SLIM bitcell array, huge Energy Delay Product (EDP) savings of ≈75× for 1T-1R (≈40× for 2T-1R) SLIM bitcell were observed for edge-detection application while EDP savings of ≈3.5× for 1T-1R (≈1.6× for 2T-1R) SLIM bitcell were observed for BNN-MLP application respectively, in comparison to conventional computing. EDP savings owing to reduction in data transfer between CPU ↔ memory is observed to be ≈780× (for both SLIM bitcells).
| S-EPMC7018944 | biostudies-literature
Project description:Dense crossbar arrays of non-volatile memory (NVM) can potentially enable massively parallel and highly energy-efficient neuromorphic computing systems. The key requirements for the NVM elements are continuous (analog-like) conductance tuning capability and switching symmetry with acceptable noise levels. However, most NVM devices show non-linear and asymmetric switching behaviors. Such non-linear behaviors render separation of signal and noise extremely difficult with conventional characterization techniques. In this study, we establish a practical methodology based on Gaussian process regression to address this issue. The methodology is agnostic to switching mechanisms and applicable to various NVM devices. We show tradeoff between switching symmetry and signal-to-noise ratio for HfO2-based resistive random access memory. Then, we characterize 1000 phase-change memory devices based on Ge2Sb2Te5 and separate total variability into device-to-device variability and inherent randomness from individual devices. These results highlight the usefulness of our methodology to realize ideal NVM devices for neuromorphic computing.
| S-EPMC5974407 | biostudies-literature
Project description:Cell-free bioproduction systems represent a promising alternative to classical microbial fermentation processes to synthesize value-added products from biological feedstocks. An essential step for establishing cell-free production systems is the identification of suitable metabolic modules with defined properties. Here we present MEMO, a novel computational approach to find smallest metabolic modules with specified stoichiometric and thermodynamic constraints supporting the design of cell-free systems in various regards. In particular, one key challenge for a sustained operation of cell-free systems is the regeneration of utilized cofactors (such as ATP and NAD(P)H). Given a production pathway with certain cofactor requirements, MEMO can be used to compute smallest regeneration modules that recover these cofactors with required stoichiometries. MEMO incorporates the stoichiometric and thermodynamic constraints in a single mixed-integer linear program, which can then be solved to find smallest suitable modules from a given reaction database. We illustrate the applicability of MEMO by calculating regeneration modules for the recently published synthetic CETCH cycle for in vitro carbon dioxide fixation. We demonstrate that MEMO is very flexible in taking into account the diverse constraints of the CETCH cycle (e.g., regeneration of 1 ATP, 4 NADPH and of 1 acetyl-group without net production of CO2 and with permitted side production of malate) and is able to determine multiple solutions in reasonable time in two large reaction databases (MetaCyc and BiGG). The most promising regeneration modules found utilize glycerol as substrate and require only 8 enzymatic steps. It is also shown that some of these modules are robust against spontaneous loss of cofactors (e.g., oxidation of NAD(P)H or hydrolysis of ATP). Furthermore, we demonstrate that MEMO can also find cell-free production systems with integrated product synthesis and cofactor regeneration. Overall, MEMO provides a powerful method for finding metabolic modules and for designing cell-free production systems as one particular application.
| S-EPMC7308072 | biostudies-literature
Project description:BackgroundCurrently, most biomedical models exist in isolation. It is often difficult to reuse or integrate models or their components, in part because they are not modular. Modular components allow the modeler to think more deeply about the role of the model and to more completely address a modeling project's requirements. In particular, modularity facilitates component reuse and model integration for models with different use cases, including the ability to exchange modules during or between simulations. The heterogeneous nature of biology and vast range of wet-lab experimental platforms call for modular models designed to satisfy a variety of use cases. We argue that software analogs of biological mechanisms are reasonable candidates for modularization. Biomimetic software mechanisms comprised of physiomimetic mechanism modules offer benefits that are unique or especially important to multi-scale, biomedical modeling and simulation.ResultsWe present a general, scientific method of modularizing mechanisms into reusable software components that we call physiomimetic mechanism modules (PMMs). PMMs utilize parametric containers that partition and expose state information into physiologically meaningful groupings. To demonstrate, we modularize four pharmacodynamic response mechanisms adapted from an in silico liver (ISL). We verified the modularization process by showing that drug clearance results from in silico experiments are identical before and after modularization. The modularized ISL achieves validation targets drawn from propranolol outflow profile data. In addition, an in silico hepatocyte culture (ISHC) is created. The ISHC uses the same PMMs and required no refactoring. The ISHC achieves validation targets drawn from propranolol intrinsic clearance data exhibiting considerable between-lab variability. The data used as validation targets for PMMs originate from both in vitro to in vivo experiments exhibiting large fold differences in time scale.ConclusionsThis report demonstrates the feasibility of PMMs and their usefulness across multiple model use cases. The pharmacodynamic response module developed here is robust to changes in model context and flexible in its ability to achieve validation targets in the face of considerable experimental uncertainty. Adopting the modularization methods presented here is expected to facilitate model reuse and integration, thereby accelerating the pace of biomedical research.
| S-EPMC4236728 | biostudies-literature
Project description:The lack of a scalable nanoparticle-based computing architecture severely limits the potential and use of nanoparticles for manipulating and processing information with molecular computing schemes. Inspired by the von Neumann architecture (VNA), in which multiple programs can be operated without restructuring the computer, we realized the nanoparticle-based VNA (NVNA) on a lipid chip for multiple executions of arbitrary molecular logic operations in the single chip without refabrication. In this system, nanoparticles on a lipid chip function as the hardware that features memory, processors, and output units, and DNA strands are used as the software to provide molecular instructions for the facile programming of logic circuits. NVNA enables a group of nanoparticles to form a feed-forward neural network, a perceptron, which implements functionally complete Boolean logic operations, and provides a programmable, resettable, scalable computing architecture and circuit board to form nanoparticle neural networks and make logical decisions.
| S-EPMC7449691 | biostudies-literature