Efficient Implementation Research Articles

Fast and Energy-Efficient Oblique Decision Tree Implementation with Potential Error Detection

In the contemporary landscape, with the proliferation of cyber-physical systems and the Internet of Things, intelligent embedded systems have become ubiquitous. These systems derive their intelligence from machine learning algorithms that are integrated within them. Among many machine learning algorithms, decision trees are often favored for implementation in such systems due to their simplicity and commendable classification performance. In this regard, we have proposed the efficient implementations of a fixed-point decision tree tailored for embedded systems. The proposed approach begins by identifying an input vector that might be classified differently by a fixed-point decision tree than by a floating-point decision tree. Upon identification, an error flag is activated, signaling a potential misclassification. This flag serves to bypass or disable the subsequent classification procedures for the identified input vector, thereby conserving energy and reducing classification latency. Subsequently, the input vector is alternatively classified based on class probabilities gathered during the training phase. In comparison with traditional fixed-point implementations, our proposed approach is proven to be 23.9% faster in terms of classification speed, consuming 11.5% less energy without compromising classification accuracy. The proposed implementation, if adopted in a smart embedded device, can provide a more responsive service to its users as well as longer battery life.

METS-R SIM: A simulator for Multi-modal Energy-optimal Trip Scheduling in Real-time with shared autonomous electric vehicles

We develop an agent-based simulator named METS-R SIM to support operational decisions for multi-modal shared autonomous vehicle (SAEV) services. Compared with existing traffic simulators, METS-R SIM offers several valuable features including: 1) A microscopic vehicle movement model for SAEV services, which allows us to explicitly model vehicular interactions and generate detailed speed and acceleration profiles for energy estimation. 2) An efficient implementation in which parallel computing is embedded in METS-R SIM which can update the state of different agents (e.g., vehicle locations in different links) simultaneously. 3) A modular and extensible framework as the simulator is built upon an agent-based modeling environment named Repast Simphony which is featured by its well-factored abstractions; in addition, a server-client structure is introduced to implement real-time operational algorithms such as energy-efficient routing and adaptive transit scheduling. 4) Open-source, reproducible with web-based visualization (METS-R SIM introduces these features to promote transparency). We validate METS-R SIM by matching the aggregated travel time and travel distance with the real observed ones obtained from New York City (NYC). We also compare the generated speed profiles qualitatively to the ones reported in published studies. We demonstrate the functionalities of our simulator by simulating SAEV services deployed to serve travel needs related to three main transportation hubs in NYC.

Coupled Ferroelectric-Photonic Memory in a Retinomorphic Hardware for In-Sensor Computing.

The development of all-in-one devices for artificial visual systems offers an attractive solution in terms of energy efficiency and real-time processing speed. In recent years, the proliferation of smart sensors in the growth of Internet-of-Things (IoT) has led to the increasing importance of in-sensor computing technology, which places computational power at the edge of the data-flow architecture. In this study, a prototype visual sensor inspired by the human retina is proposed, which integrates ferroelectricity and photosensitivity in two-dimensional (2D) α-In2Se3 material. This device mimics the functions of photoreceptors and amacrine cells in the retina, performing optical reception and memory computation functions through the use of electrical switching polarization in the channel. The gate-tunable linearity of excitatory and inhibitory functions in photon-induced short-term plasticity enables to encode and classify 12000 images in the Mixed National Institute of Standards and Technology (MNIST) dataset with remarkable accuracy, achieving ≈94%. Additionally, in-sensor convolution image processing through a network of phototransistors, with five convolutional kernels electrically pre-programmed into the transistors is demonstrated. The convoluted photocurrent matrices undergo straightforward arithmetic calculations to produce edge and feature-enhanced scenarios. The findings demonstrate the potential of ferroelectric α-In2Se3 for highly compact and efficient retinomorphic hardware implementation, regardless of ambipolar transport in the channel.

Multi-scale assessment of global warming mitigation potential of anaerobic digestion for food waste management in the United States: A comparison of Life Cycle Assessment approaches

This study provides a holistic Life Cycle Assessment (LCA) for evaluating the decarbonization potential of using Anaerobic Digestion (AD) in food waste treatment. The LCA is based on the development of a new flexible tool that allows the integration of the LCA at the process level with geospatial data and therefore includes a wide range of factors such as multiple feedstocks, technologies for end-use applications, and facility locations. This tool estimates a potential for reducing 21.8 MMtons CO2eq/y (345 gCO2eq/kgwaste) and generating 161 TBTU/y of renewable natural gas for the whole United States by managing food waste based on waste-hierarchy. This study also identifies those states with a more efficient AD implementation potential regarding CO2eq avoidance. It highlights the importance of including region-specific conditions since feedstock diversion depends on the state and is the main contributor to the decarbonization potential, up to 60 % of the avoided emissions.

A Star Network of Bipolar Memristive Devices Enables Sensing and Temporal Computing.

Temporal (race) computing schemes rely on temporal memories, where information is represented with the timing of signal edges. Standard digital circuit techniques can be used to capture the relative timing characteristics of signal edges. However, the properties of emerging device technologies could be particularly exploited for more efficient circuit implementations. Specifically, the collective dynamics of networks of memristive devices could be leveraged to facilitate time-domain computations in emerging memristive memories. To this end, this work studies the star interconnect configuration of bipolar memristive devices. Through circuit simulations using a behavioral model of voltage-controlled bipolar memristive devices, we demonstrated the suitability of such circuits in two different contexts, namely sensing and "rank-order" coding. We particularly analyzed the conditions that the employed memristive devices should meet to guarantee the expected operation of the circuit and the possible effects of device variability in the storage and the reproduction of the information in arriving signal edges. The simulation results in LTSpice validate the correct operation and confirm the promising application prospects of such simple circuit structures, which, we show, natively exist in the crossbar geometry. Therefore, the star interconnect configuration could be considered for temporal computations inside resistive memory (ReRAM) arrays.

Modelling stellar variability in archival HARPS data: I - Rotation and activity properties with multidimensional Gaussian processes

ABSTRACT Although instruments for measuring the radial velocities (RVs) of stars now routinely reach sub-metre per second accuracy, the detection of low-mass planets is still very challenging. The rotational modulation and evolution of spots and/or faculae can induce variations in the RVs at the level of a few m s–1 in Sun-like stars. To overcome this, a multidimensional Gaussian Process framework has been developed to model the stellar activity signal using spectroscopic activity indicators together with the RVs. A recently published computationally efficient implementation of this framework, S + LEAF 2, enables the rapid analysis of large samples of targets with sizeable data sets. In this work, we apply this framework to HARPS observations of 268 well-observed targets with precisely determined stellar parameters. Our long-term goal is to quantify the effectiveness of this framework to model and mitigate activity signals for stars of different spectral types and activity levels. In this first paper in the series, we initially focus on the activity indicators (S-index and Bisector Inverse Slope), and use them to (a) measure rotation periods for 49 slow rotators in our sample, (b) explore the impact of these results on the spin-down of middle-aged late F, G, and K stars, and (c) explore indirectly how the spot to facular ratio varies across our sample. Our results should provide valuable clues for planning future RV planet surveys such as the Terra Hunting Experiment or the PLATO ground-based follow-up observations programme, and help fine-tune current stellar structure and evolution models.

Efficient Version Space Algorithms for Human-in-the-loop Model Development

When active learning (AL) is applied to help users develop a model on a large dataset through interactively presenting data instances for labeling, existing AL techniques often suffer from two main drawbacks: First, to reach high accuracy they may require the user to label hundreds of data instances, which is an onerous task for the user. Second, retrieving the next instance to label from a large dataset can be time-consuming, making it incompatible with the interactive nature of the human exploration process. To address these issues, we introduce a novel version-space-based active learner for kernel classifiers, which possesses strong theoretical guarantees on performance and efficient implementation in time and space. In addition, by leveraging additional insights obtained in the user labeling process, we can factorize the version space to perform active learning in a set of subspaces, which further reduces the user labeling effort. Evaluation results show that our algorithms significantly outperform state-of-the-art version space strategies, as well as a recent factorization-aware algorithm, for model development over large datasets.

GPU Algorithms for Structured Sparse Matrix Multiplication with Diagonal Storage Schemes

Matrix–matrix multiplication is of singular importance in linear algebra operations with a multitude of applications in scientific and engineering computing. Data structures for storing matrix elements are designed to minimize overhead information as well as to optimize the operation count. In this study, we utilize the notion of the compact diagonal storage method (CDM), which builds upon the previously developed diagonal storage—an orientation-independent uniform scheme to store the nonzero elements of a range of matrices. This study exploits both these storage schemes and presents efficient GPU-accelerated parallel implementations of matrix multiplication when the input matrices are banded and/or structured sparse. We exploit the data layouts in the diagonal storage schemes to expose a substantial amount of fine-grained parallelism and effectively utilize the GPU shared memory to improve the locality of data access for numerical calculations. Results from an extensive set of numerical experiments with the aforementioned types of matrices demonstrate orders-of-magnitude speedups compared with the sequential performance.

Massively parallel implementation of gradients within the random phase approximation: Application to the polymorphs of benzene.

The Random-Phase approximation (RPA) provides an appealing framework for semi-local density functional theory. In its Resolution-of-the-Identity (RI) approach, it is a very accurate and more cost-effective method than most other wavefunction-based correlation methods. For widespread applications, efficient implementations of nuclear gradients for structure optimizations and data sampling of machine learning approaches are required. We report a well scaling implementation of RI-RPA nuclear gradients on massively parallel computers. The approach is applied to two polymorphs of the benzene crystal obtaining very good cohesive and relative energies. Different correction and extrapolation schemes are investigated for further improvement of the results and estimations of error bars.

Divide-and-conquer-based RDO-free CU Partitioning for 8K Video Compression

8K (7689×4320) ultra-high definition (UHD) videos are growing popular with the improvement of human visual experience demand. Therefore, the compression of 8K UHD videos has become a top priority in the third-generation audio video coding standard (AVS3). However, as an important part of the coding standard promotion, the real-time hardware implementation for AVS3-based 8K UHD video intra coding is severely hindered, especially in the coding unit (CU) partition stage. To break through the limitation, this article proposes a divide-and-conquer-based rate-distortion-optimization-free (RDO-free) CU partitioning algorithm for efficient hardware implementation. Aimed at the complex CU partition in AVS3, we separately design a lightweight optimization for original partitioning rules to improve division efficiency and a decision tree-based RDO-free CU decision framework to eliminate the latency caused by the waiting for rate-distortion cost calculation in RDO strategy. Afterward, a divide-and-conquer-based hardware-friendly gradient difference calculating approach is devised to accelerate the learning feature extracting speed. To ensure that the proposed algorithm is sufficient to support the real-time CU partition for 8K videos, we also develop a hardware architecture based on FPGA. Experimental results illustrate that the software coding performance of our algorithm is significantly ahead of the efficient implementation uAVS3e for AVS3 and the reference software HM-16.20 for High Efficiency Video Coding (HEVC), even though there is 9.96% loss on BD-Rate Y. Considering its importance for the hardware implementation of AVS3-based 8K real-time encoder, the coding loss is acceptable. Moreover, the hardware simulation results on VU440 FPGA with Vivado 2019 show that our algorithm can support 61.12 frames per second (fps) CU partition for 8K UHD videos with only 0.00%, 0.00%, 1.01%, and 7.78% consumption of BRAM_18K, DSP48E, FF, and LUT, respectively. Additionally, with dual-path parallelism, 122.24 fps also can be implemented with controllable resource utilization, which achieves the state-of-the-art performance.

Efficient hybrid oversampling and intelligent undersampling for imbalanced big data classification

Imbalanced classification is a well-known challenge faced by many real-world applications. This issue occurs when the distribution of the target variable is skewed, leading to a prediction bias toward the majority class. With the arrival of the Big Data era, there is a pressing need for efficient solutions to solve this problem. In this work, we present a novel resampling method called SMOTENN that combines intelligent undersampling and oversampling using a MapReduce framework. Both procedures are performed on the same pass over the data, conferring efficiency to the technique. The SMOTENN method is complemented with an efficient implementation of the neighborhoods related to the minority samples. Our experimental results show the virtues of this approach, outperforming alternative resampling techniques for small- and medium-sized datasets while achieving positive results on large datasets with reduced running times.

Efficient single-run implementation of generalized Einstein relation to compute transport coefficients: A binary-based time sampling.

A robust and simple implementation of the generalized Einstein formulation using single equilibrium molecular dynamics simulation is introduced to compute diffusion and shear viscosity. The unique features underlying this framework are as follows: (1) The use of a simple binary-based method to sample time-dependent transport coefficients results in a uniform distribution of data on a logarithmic time scale. Although we sample "on-the-fly," the algorithm is readily applicable for post-processing analysis. Overlapping same-length segments are not sampled as they indicate strong correlations. (2) Transport coefficients are estimated using a power law fitting function, a generalization of the standard linear relation, that accurately describes the long-time plateau. (3) The use of a generalized least squares (GLS) fitting estimator to explicitly consider correlations between fitted data points results in a reliable estimate of the statistical uncertainties in a single run. (4) The covariance matrix for the GLS method is estimated analytically using the Wiener process statistics and computed variances. (5) We provide a Python script to perform the fits and automate the procedure to determine the optimal fitting domain. The framework is applied to two fluids, binary hard sphere and a Lennard-Jones near the triple point, and the validity of the single-run estimates is verified against multiple independent runs. The approach should be applicable to other transport coefficients since the diffusive limit is universal to all of them. Given its rigor and simplicity, this methodology can be readily incorporated into standard molecular dynamics packages using on-the-fly or post-processing analysis.

Linearithmic and unbiased implementation of DeLong’s algorithm for comparing the areas under correlated ROC curves

Receiver operating characteristic (ROC) analysis is a widespread framework in two-group classification problems. The area under the ROC curve (AUC) has been a popular figure of merit to summarize the performance of binary classifiers. Because of its simplicity of implementation, DeLong’s algorithm (DLA) has been widely used in practice. However, DLA suffers two drawbacks of time inefficiency and biasedness. The authors’ previous work has overcome the first drawback. Specifically, the time complexity of DLA is reduced from the quadratic order down to a linearithmic one, by exploiting the relationship between the Heaviside function and the mid-ranks of samples. Unfortunately, the second drawback of DLA remains unaddressed, to the best of our knowledge. Motivated by such unsatisfactory situation, in this paper, the authors correct the bias of DLA with the assistance of an efficient implementation of Kendall’s tau (KT). The overall computational load of the improved DLA (being strictly unbiased) is in linearithmic time. Theoretical analyses and numerical results verified the proposed algorithmic findings in this work.

A dual-modality optical system for single-pixel imaging and transmission through scattering media

It is well recognized that it is difficult to develop an optical system to retrieve effective information when dynamic and turbid water exists in an optical channel. It could be more challenging to incorporate dual or multiple modalities in one optical system. In this Letter, we report a dual-modality optical system for single-pixel imaging (SPI) and transmission through scattering media. A series of mutually-orthogonal random illumination patterns are designed and adopted to realize high-resolution image recovery in SPI. The data to be transmitted are also encoded into random illumination patterns in a differential way, and high-fidelity free-space optical data transmission can be simultaneously realized. Experimental results validate feasibility of the proposed optical system and its high robustness against scattering. The developed dual-modality optical system realizes high-resolution SPI and high-fidelity data transmission in scattering media using only one set of realizations, offering an efficient implementation with reduced power and equipment requirements. The proposed method is promising toward the development of an integrated system with multiple modalities for optical information retrieval, especially in dynamic scattering media.

A Two‐Step Entanglement Concentration for the Less‐Entangled Four‐Photon Cluster States

AbstractEntanglement concentration serves as an essential tool to obtain the maximally entangled states from the less‐entangled states in quantum information processing tasks. An efficient two‐step entanglement concentration protocol (ECP) is proposed to concentrate the less‐entangled four‐photon cluster states with four unknown coefficients into the maximally entangled states in this paper. The protocol utilizes parity‐check gate and photon‐polarized controlled‐Z gate, constructed with a single semiconductor quantum dot in a single optical microcavity (quantum dot‐cavity coupled system) and several linear optical elements. The discarded quantum states can be recycled in the next round concentration to achieve a higher success probability. Under conditions of strong coupling and weak side leakage rate, the ECP can tend toward ideal fidelity and efficiency. Compared to the previously proposed ECPs that utilize various auxiliary tools such as additional entangled states and complex quantum logic gates, the ECP significantly simplifies the analysis process and enables efficient implementation. Thus, the ECP is more economical and feasible in future optical systems.

Entropy and the Kullback–Leibler Divergence for Bayesian Networks: Computational Complexity and Efficient Implementation

Bayesian networks (BNs) are a foundational model in machine learning and causal inference. Their graphical structure can handle high-dimensional problems, divide them into a sparse collection of smaller ones, underlies Judea Pearl’s causality, and determines their explainability and interpretability. Despite their popularity, there are almost no resources in the literature on how to compute Shannon’s entropy and the Kullback–Leibler (KL) divergence for BNs under their most common distributional assumptions. In this paper, we provide computationally efficient algorithms for both by leveraging BNs’ graphical structure, and we illustrate them with a complete set of numerical examples. In the process, we show it is possible to reduce the computational complexity of KL from cubic to quadratic for Gaussian BNs.

An efficient sequential consistency implementation with dynamic race detection for GPUs

As GPUs are being used for general purpose computations, applications with different memory access requirements have emerged. In spite of the growing demand, only few GPU coherence protocols and memory models have been explored in research, and even fewer models have been implemented in products. However, in the CPU domain a diverse range of memory models for parallel programming have been proposed, which explore the interplay between performance and programmability.Sequential consistency (SC) is one of the strict memory models. It provides the most programmer intuitive execution of memory operation but it imposes strict ordering restrictions on memory operations that cause performance overhead. Hence, implementing and supporting SC is one of the most challenging tasks in any computing platform, and GPUs are no exception. As such in this paper, we propose a GPU architecture that implements SC memory model with minimal performance and power overhead. We achieve this goal by designing a mechanism to detect races between different streaming multiprocessors (SMs) dynamically at runtime. The race is detected using a signature-based mechanism to keep track of sets of unseen updates for each SM which significantly reduces the hardware implementation cost, with a small increase in invalidation traffic. Our experiments show that dynamic race detection can be used to implement sequential consistency with 5% performance overhead.

Looking for winning policies to address the climate issue in EU-aviation

This paper analyzes the different climate policies that are used by the EU in the aviation sector. Using a simple model, we study their effects on welfare as well as on their acceptability. The efficiency and acceptability depend on the strength of the international agreements (Paris, 2015; CORSIA, 2020) and on the implementation of the policy within the EU. Four different policies are discussed: carbon taxes, tradeable emission permits, Sustainable Aviation Fuels (SAF) blending mandates and fuel efficiency standards for new aircrafts. The Paris and the CORSIA agreements are both weak agreements where, in the absence of sanctions, the implementation depends on the will of the political majority of the moment. In terms of implementation, it is shown that for intra EU aviation, the insertion of aviation in the EU-ETS with partially auctioned emission permits is probably the most robust and efficient implementation of the Paris agreement as it can result in net zero emissions in 2040–2050. Subsidies for R&D in fuel efficiency and new fuels would be good complements to the ETS system as they also have a positive spillover for the non-EU aviation via better aircrafts and better fuels. SAF mandates are likely to be inefficient implementations of climate policies, they exist because they are more acceptable for the airlines and intensive air transport users.

Temporal data modelling and integrity constraints in relational databases*

There are key issues in temporal relational data modelling such as representation of temporal data, temporal grouping identifiers, primary keys of temporal relations and temporal integrity constraints. Temporal grouping is a fundamental aspect of temporal data modelling and is the basis for the specification and enforcement of temporal integrity constraints. We extend and generalize traditional integrity constraints for the temporal domain by formally defining the temporal grouping identifier and show how it can replace a temporal primary key like a temporal atom, a temporal integrity constraint has a data and a temporal component. We also introduce the notion of consistent evolution of a temporal database that lays the foundation for an efficient implementation of temporal integrity constraints. We also provide clauses that need to be added to SQL for specifying the temporal integrity constraints.

PROFIT: projection-based test in longitudinal functional data

In many modern applications, a dependent functional response is observed for each subject over repeated time, leading to longitudinal functional data. In this paper, we propose a novel statistical procedure to test whether the mean function varies over time. Our approach relies on reducing the dimension of the response using data-driven orthogonal projections, and employs likelihood-based hypothesis testing. We investigate the methodology theoretically and discuss a computationally efficient implementation. The proposed test maintains the Type-1 error rate, and shows excellent power to detect departures from the null hypothesis in finite sample simulation studies. We apply our method to the longitudinal diffusion tensor imaging study of multiple sclerosis (MS) patients to formally assess whether the brain's healthy tissue, as summarised by the fractional anisotropy (FA) profile, degrades over time during the study period.