Memory Layer Research Articles

Time Series Storage Method of Multi-Valued Attribute Data in Energy Big Data Center

Herein, the time series storage method of multi-valued attribute data, aimed at improving the efficiency of writing and querying data in “energy” big data centers is reported. Through rule construction and rule iteration of feature sequence, the time series features of multi-valued attribute data are extracted, and the “component attribute nearest neighbor propagation method” is used for clustering; the data are divided into cold, warm, and hot data. A storage engine with a solid state disk layer, mechanical hard disk layer, and memory layer has been designed, and the efficiency of data writing and query through migration operation is improved. The experimental results demonstrate that this method can effectively extract the time series features of multi-valued attribute data, and the Pre value of the clustering time series is higher than 0.93, which effectively improves the data writing and query efficiency of the energy big data center.

A deep hybrid network for significant wave height estimation

The influence of weather conditions on sea state, and in particular on the dynamic evolution of waves, is an important issue that affects several areas, including maritime traffic and the planning of coastal works. To collect relevant data, buoys are used to set up distributed sensor networks along coastal areas. However, unfavourable weather conditions can lead to downtime, which can be extended due to maintenance issues. The ability to improve the robustness of these sensor systems using predictive models, i.e. digital twins, to interpolate and extrapolate missing data is an important and growing area of research. To accomplish such a task, models must be found that can account for both the spatial and temporal dynamics of the input data to correctly estimate the variables of interest. In this work, a deep learning architecture is proposed to realize a digital twin for the monitoring buoy for significant wave height estimation using spatial and temporal information about the wind field in the area of interest. The proposed methodology was applied to a case study using wave height data from an Italian Sea Monitoring Network buoy installed near the coast of Sicily and wind field data from the Copernicus Climate Change Service ERA5 reanalysis. The reported results show that the use of a multi-block hybrid deep neural network consisting of convolutional layers for spatial feature extraction and short-term memory layers for modelling the involved dynamics, which takes into account the buoy surrounding area, outperforms other empirical, numerical, machine learning and deep learning methods used in the literature.

Forecasting price in a new hybrid neural network model with machine learning

A key aspect of asset investment and risk management is the study of forecasting stock prices. We investigate the machine learning stock price prediction in a new hybrid neural network model and put forth a forecasting method based on machine learning, composite data preprocessing method and the proposed new neural network model. To address the challenge of predicting stock prices in the face of market complexity and noise, we use the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) algorithm and the Savitzky–Golay (SG) filter to de-noise and enhance the data, and employ a neural network with three convolutional layers and a long short-term memory (LSTM) layer that enables it to capture complex temporal patterns in the data. We propose a new hybrid neural network prediction model (CEEMDAN-S-C-LSTM) and adopt a machine learning approach to compare it with the benchmark model using CSI 300 index data. The empirical results validate the effectiveness of the frequency decomposition algorithm and the convolutional layer, and demonstrate that our proposed model outperforms the benchmark model. Compared to the best benchmark model, the CEEMDAN-S-C-LSTM model proposed in this study demonstrates a significant improvement in performance. Specifically, it shows a 45.33% reduction in mean absolute error (MAE), 43.44% reduction in root mean square error (RMSE), 45.01% reduction in mean absolute percentage error (MAPE), and 3.90% improvement in coefficient of determination (R2). The study also explores the effect of different numbers of convolution layers and SG filters on the hybrid model. Our research expands on the use of neural networks and machine learning, offering a novel technical approach to making investment decisions and managing risks in financial systems.

Prediction of instantaneous flow characteristics of hydrocyclone with long short-term memory network based on computational fluid dynamics data

Accelerating the prediction time of separation performance and flow field characteristics in industrial hydrocyclones holds paramount importance for real-time control. Machine learning methods exhibit significant advantages in this particular aspect. This study presents a network model that integrates a long short-term memory (LSTM) layer with a fully connected layer to accurately forecast the flow field and separation performance under various operational conditions, leveraging CFD data. The study evaluates the effect of factors such as the number of LSTM layers and neurons per layer on the model's performance, aiming to identify the optimal network structure. The results demonstrate that the predicted flow field and performance of the hydrocyclones using LSTM closely align with the outcomes from CFD simulations. The average absolute percentage error is constrained to within 6%, which significantly enhances the prediction efficiency. The findings contribute to a more efficient and automated separation process in industrial environments.

A Flexible Nickel-Oxide-Based RRAM Device Prepared Using the Solution Combustion Method

Binary metal oxide materials, such as nickel oxide, are widely used in flexible resistive variable memory devices due to advantages such as their easily controllable material composition, simple structural composition, and good compatibility between manufacturing processes and complementary metal oxide processes. In this work, a solution combustion method was employed to prepare NiOx thin films for use as a resistive switching layer of resistance random-access memory. The formation temperature of the NiOx layer in the RRAM device was kept as low as 175 °C, making the device compatible with flexible substrates. With polyethylene naphthalenediate as the substrate, the NiOx-based RRAM exhibited obvious bipolar resistance-switching properties, superb bending resistance, and good stability. Through theoretical fitting and structural characterization, the conduction mechanisms were attributed to the combination of the space-charge-limited current and Ohmic mechanisms, while the valence change mechanism was determined to be the resistive switching mechanism. This study demonstrates a low-temperature and scalable approach to constructing NiOx-based RRAM devices on flexible substrates.

State of charge estimation for Li-ion battery during dynamic driving process based on dual-channel deep learning methods and conditional judgement

The accurate estimation of the state of charge (SOC) is crucial for the safe and reliable operation of Li-ion batteries. In this paper, novel dual-channel deep learning methods were proposed for SOC estimation during dynamic driving cycles. For dual-channel deep learning methods, each channel of the deep learning methods could be a single convolutional neural network (CNN) layer, single long short-term memory (LSTM) layer or sequentially combined structure of CNN and LSTM. After evaluation, it could be found that dual-channel deep learning methods could achieve higher accuracy than conventional single-channel method, however, the model size and floating-point operations (FLOPs) could be either reduced or increased. Among all the dual-channel deep learning methods, CNN + LSTM-CNN + LSTM method could achieve the highest accuracy and the shortest training time with sacrifices in computational complexity. CNN–CNN + LSTM method could achieve higher accuracy than the single-channel deep learning method with reduced computational complexity and model size. As the error of SOC estimation is highly correlated with voltage signals, conditional judgement was then integrated with dual-channel deep learning method to obtain higher accuracy. After integrated with conditional judgement, the averaged error of SOC estimation could be reduced by more than 45% without a substantial sacrifice in computational time.

Computerized Segmentation Method for Nonmasses on Breast DCE-MRI Images Using ResUNet++ with Slice Sequence Learning and Cross-Phase Convolution

The purpose of this study was to develop a computerized segmentation method for nonmasses using ResUNet++ with a slice sequence learning and cross-phase convolution to analyze temporal information in breast dynamic contrast material-enhanced magnetic resonance imaging (DCE-MRI) images. The dataset consisted of a series of DCE-MRI examinations from 54 patients, each containing three-phase images, which included one image that was acquired before contrast injection and two images that were acquired after contrast injection. In the proposed method, the region of interest (ROI) slice images are first extracted from each phase image. The slice images at the same position in each ROI are stacked to generate a three-dimensional (3D) tensor. A cross-phase convolution generates feature maps with the 3D tensor to incorporate the temporal information. Subsequently, the feature maps are used as the input layers for ResUNet++. New feature maps are extracted from the input data using the ResUNet++ encoders, following which the nonmass regions are segmented by a decoder. A convolutional long short-term memory layer is introduced into the decoder to analyze a sequence of slice images. When using the proposed method, the average detection accuracy of nonmasses, number of false positives, Jaccard coefficient, Dice similarity coefficient, positive predictive value, and sensitivity were 90.5%, 1.91, 0.563, 0.712, 0.714, and 0.727, respectively, larger than those obtained using 3D U-Net, V-Net, and nnFormer. The proposed method achieves high detection and shape accuracies and will be useful in differential diagnoses of nonmasses.

Tetherless and Batteryless Soft Navigators and Grippers.

Remotely controllable soft actuators have promising potential applications in many fields including soft robotics, exploration, and invasion medical treatment. Shape memory polymers could store and release energy, resulting in shape deformation, and have been regarded as promising candidates to fabricate untethered soft robots. Herein, an untethered and battery-free soft navigator and gripper based on a shape memory hydrogel is presented. The shape memory hydrogel is obtained through hydrogen bonding between gelatin and tannic acid, and the hydrogel displays excellent shape memory properties on the basis of hydrogen bonding and the coil-triple helix transition of gelatin. Moreover, Fe3O4 nanoparticles are introduced to endow the hydrogel magnetic responsiveness and photothermal conversion capacity. Finally, the shape memory hydrogel in a stretched state is assembled with an inert hydrogel to achieve a bilayer hydrogel actuator, which could produce complex shape transformation due to the shape recovery of the shape memory layer induced by heat or light. Taking advantage of the magnetically control and light-responsive shape deformation, remotely controllable soft grippers that could navigate through tortuous paths and grasp objects from a hard-to-reach place have been accomplished. This approach will inspire the design and fabrication of novel shape memory hydrogels as remotely controllable soft robots.

Monuments of Italian Colonialism as a Transcultural Heritage

This essay addresses the material legacies of Italian colonialism, reframing them as a transcultural heritage. With a focus on the complex interplay between memory and displacement, the study retraces the biographies of selected historical monuments from the late nineteenth century to fascism up until postcolonial Italy. Comparing the life and afterlives of the Monument to the Fallen Soldiers of Dogali, still preserved in a public space in Rome, with monuments implicated in processes of restitution, such as the Axum Stele and the statue of the Lion of Judah, both returned to Ethiopia, the essay reflects upon the dynamics of invisibility and absence, examining decolonial claims, and new layers of memory that currently overlap in politically charged spaces.

Wine quality assessment through lightweight deep learning: integrating 1D-CNN and LSTM for analyzing electronic nose VOCs signals

PurposeIn the wine industry, maintaining superior quality standards is crucial to meet the expectations of both producers and consumers. Traditional approaches to assessing wine quality involve labor-intensive processes and rely on the expertise of connoisseurs proficient in identifying taste profiles and key quality factors. In this research, we introduce an innovative and efficient approach centered on the analysis of volatile organic compounds (VOCs) signals using an electronic nose, thereby empowering nonexperts to accurately assess wine quality.Design/methodology/approachTo devise an optimal algorithm for this purpose, we conducted four computational experiments, culminating in the development of a specialized deep learning network. This network seamlessly integrates 1D-convolutional and long-short-term memory layers, tailor-made for the intricate task at hand. Rigorous validation ensued, employing a leave-one-out cross-validation methodology to scrutinize the efficacy of our design.FindingsThe outcomes of these e-demonstrates were subjected to meticulous evaluation and analysis, which unequivocally demonstrate that our proposed architecture consistently attains promising recognition accuracies, ranging impressively from 87.8% to an astonishing 99.41%. All this is achieved within a remarkably brief timeframe of a mere 4 seconds. These compelling findings have far-reaching implications, promising to revolutionize the assessment and tracking of wine quality, ultimately affording substantial benefits to the wine industry and all its stakeholders, with a particular focus on the critical aspect of VOCs signal analysis.Originality/valueThis research has not been published anywhere else.

Traffic Congestion Prediction Using Feature Series LSTM Neural Network and a New Congestion Index

Large and expanding cities suffer from a traffic congestion problem that harms the environment, travelers, and the economy. This paper aims to predict short term traffic congestion on a road section of expressway in Delhi city. For this purpose, we first propose a traffic congestion index based on traffic speed and flow. Clustering techniques and the Greenshield’s model were used for the derivation of the congestion index. Using this congestion index, congested time intervals of each day and each location of a weekday were identified. This study also introduces a feature series long short-term memory neural network (FSLSTMNN), which links a long short-term memory (LSTM) layer to each feature. It is trained using the many heterogeneous traffic features data collected in Delhi city for the next five minutes of traffic flow and speed prediction. FSLSTMNN achieved the good capability to learn feature series data. We also trained several traditional and deep-learning models using the same traffic data. The FSLSTMNN reduces mean absolute error 12.90% and 17.13%, respectively, in speed and traffic flow prediction compared to the second good-performance long short-term memory neural network (LSTMNN). Finally, traffic congestion is predicted classwise (light, medium, and congested) using the developed congestion index and traffic speed and flow predicted by the FSLSTMNN. Predicted results are consistent with the measured field data. Study results confirm that the developed congestion index and FSLSTMNN can be used successfully to predict traffic congestion.

Using Deep Learning for the Taxonomic Classification of Microbial Sequences

Microbes are common creatures and play a crucial role in our world. Thus, the understanding of microbial communities brings benefits to human lives. Because the material samples of microbes contain sequences belonging to different organisms, an important task in analyzing processes is to classify the sequences into groups of different species or closely related organisms, called metagenomic classification. Many classification approaches were proposed to analyze the metagenomic data. However, due to the complexity of microbial samples, the accuracy performance of those methods still remains a challenge. This study applies an effective deep learning framework for the classification of microbial sequences. The proposed architecture combines a sequence embedding layer with other layers of a bidirectional Long Short-Term Memory, Seft-attention, and Dropout mechanisms for feature learning. Experimental results demonstrate the strength of the proposed method on datasets of real metagenomes.

Suppressed, Adopted and Invented Memories

The Gospel of John reflects several layers of social memory and theological creativity concerning Jesus’s death. In the early material, there seems to be a suppressed awareness of Jesus’s fate and an unwillingness to unfold it in narrative form – something that recalls the hypothetical sayings gospel Q and the Gospel of Thomas. There is also a search for alternative, figurative ways to visualize the endpoint of Jesus’s earthly life. Eventually, the narrative memory of Jesus’s passion, as told in Mark and Matthew, was adopted with some modifications. Among the modifications of the passion storyline is the narrativization of the image of Jesus as a Paschal Lamb, an image already known to Paul. The most remarkable innovation, however, was the figure of the “Beloved Disciple” as an eyewitness to Jesus’s passion and death.

Inceptionv3‐LSTM‐COV: A multi‐label framework for identifying adverse reactions to COVID medicine from chemical conformers based on Inceptionv3 and long short‐term memory

AbstractDue to the global COVID‐19 pandemic, distinct medicines have been developed for treating the coronavirus disease (COVID). However, predicting and identifying potential adverse reactions to these medicines face significant challenges in producing effective COVID medication. Accurate prediction of adverse reactions to COVID medications is crucial for ensuring patient safety and medicine success. Recent advancements in computational models used in pharmaceutical production have opened up new possibilities for detecting such adverse reactions. Due to the urgent need for effective COVID medication development, this research presents a multi‐label Inceptionv3 and long short‐term memory methodology for COVID (Inceptionv3‐LSTM‐COV) medicine development. The presented experimental evaluations were conducted using the chemical conformer image of COVID medicine. The features of the chemical conformer are denoted utilizing the RGB color channel, which is extracted using Inceptionv3, GlobalAveragePooling2D, and long short‐term memory (LSTM) layers. The results demonstrate that the efficiency of the Inceptionv3‐LSTM‐COV model outperformed the previous study's performance and achieved better results compared to MLCNN‐COV, Inceptionv3, ResNet50, MobileNetv2, VGG19, and DenseNet201 models. The proposed model reported the highest accuracy value of 99.19% in predicting adverse reactions to COVID medicine.

BERT embedding model for arabic temporal relation classification using hybrid deep learning architectures and linguistic features

This paper introduces a novel neural network architecture for classifying temporal relationships among events in Arabic sentences. Our model integrates a deep learning pipeline that combines multiple techniques. Initially, the Bidirectional Encoder Representations from Transformers (BERT) model is employed to obtain the contextual representation of each word. Furthermore, the model integrates the part-of-speech (POS) representation, the position of events, and the output from a convolutional neural network (CNN) based at the sentence’s head. Collectively, these features capture the complex relationships between words in their context. The architecture also incorporates two sequential Bidirectional Long Short-Term Memory (BiLSTM) layers. These are complemented by an attention mechanism, which assesses the significance of each word in terms of temporal relationship types. Another CNN layer processes the entire sentence, and finally, a fully connected layer with Softmax determines the temporal relation category, drawing on vectors from the BiLSTM layers, the attention mechanism, and the CNN. Experimental results on the Ara-TimeBank corpus reveal that our model achieves an 89% F1-score, outperforming prior work in this domain.

H3DM: A High-bandwidth High-capacity Hybrid 3D Memory Design for GPUs

Graphics Processing Units (GPUs) are widely used for modern applications with huge data sizes. However, the performance benefit of GPUs is limited by their memory capacity and bandwidth. Although GPU vendors improve memory capacity and bandwidth using 3D memory technology (HBM), many important workloads with terabytes of data still cannot fit in the provided capacity and are bound by the provided bandwidth. With a limited GPU memory capacity, programmers should handle the data movement between GPU and host memories by themselves, causing a significant programming burden. To improve programming ease, GPUs use a unified address space with the host that allows over-subscribing GPU memory, but this approach is not effective in terms of performance once GPUs encounter memory page faults. Many recent works have tried to remedy capacity and bandwidth bottlenecks using dense non-volatile memories (NVMs) and true-3D stacking. However, these works mainly focus on one bottleneck or do not provide a scalable solution that fits future requirements. In this paper, we investigate true-3D stacking of dense, low-power, and refresh-free non-volatile phase change memory (PCM) on top of state-of-the-art GPU configurations to provide higher capacity and bandwidth within the available area and power budget. The higher density and lower power consumption of PCM provide higher capacity through integrating more cells in each 3D layer and enabling stacking more layers. However, we observe that stacking more than six layers of pure-PCM memory violates the thermal constraint and severely harms the performance and power efficiency due to its higher write latency and energy. Further, it degrades the lifetime of GPU to less than one year. Utilizing a hybrid architecture that leverages the benefits of both DRAM and PCM memories has been widely studied by prior proposals; however, true-3D integration of such a hybrid memory architecture especially on top of state-of-the-art powerful GPU architecture has not been investigated yet. We experimentally demonstrate that by covering 80% of write requests in DRAM and eliminating refresh overhead, true-3D stacking of eight 32GB layers of PCM along with two 8GB layers of DRAM is possible resulting in a total of 272GB memory capacity. Based on the explored design requirements, We propose a 3D high-bandwidth high-capacity hybrid memory (H3DM) system utilizing a hybrid-3D (H3D)-aware remapping scheme to reduce expensive PCM writes to under 20% while avoiding DRAM refresh overhead. H3DM improves the performance up to 291% compared to the baseline GPU architecture while remaining within only 3% of an ideal case with DRAM-like access latency, on average. Moreover, by increasing the dataset size above the baseline GPU memory space, H3DM improves performance and power up to 648% and 87% compared to the baseline GPU architecture since it avoids expensive data transfers through off-chip communication links.

Intelligent fault diagnosis of storage stacking machinery under variable working conditions using attention-based adaptive multimodal feature fusion networks

Due to the harsh working environment of storage stacking machinery, the fault information of important components is significantly complex, which leads to the problem of low classification accuracy and high computational complexity of existing deep learning-based fault diagnosis methods. To alleviate the problem, this paper presents a novel architecture named attention-based adaptive multimodal feature fusion networks for intelligent fault diagnosis of storage stacking machinery, which is aimed at improving the diagnostic precision and robustness of feature fusion network and learning the broader feature representation. Firstly, the long short-term memory layer is introduced to extract the feature information of multiple time steps to improve the self-extraction ability of multi-temporal features. Then, the maximum temporal feature fusion module is utilized to highlight the recognizability of deep fusion features. Finally, a residual layer with spanning connections is added to increase the utilization and characterization capability of deep fusion features. Experimental results demonstrate the effectiveness and superiority of the proposed method in intelligent fault diagnosis of storage stacking machinery under variable working conditions compared with some state-of-the-art deep learning-based methodologies.

RNN-Based Approach for Broccoli Harvest Time Forecast

This article investigates approaches for broccoli harvest time prediction through the application of various machine learning models. This study’s experiment is conducted on a commercial farm in Ecuador, and it integrates in situ weather and broccoli growing cycle observations made over seven years. This research incorporates models such as the persistence, thermal, and calendar models, demonstrating their strengths and limitations in calculating the optimal broccoli harvest day. Additionally, Recurrent Neural Network (RNN) models with Long Short-term Memory (LSTM) layers were developed, showcasing enhanced accuracy with an error of less than 2.5 days on average when combined with outputs from the calendar model. In the final comparison, the RNN models outperformed both the thermal and calendar models, with an error of 3.14 and 2.5 days, respectively. Furthermore, this article explores the impact of utilizing Global Ensemble Forecast System forecast weather data as a supplementary source to the in situ observations on model accuracy. The analysis revealed the limited effect of extension with a 9-day forecast on the experimental field, reaching an error reduction of up to 0.04 days. The findings provide insights into the effectiveness of different modeling approaches for optimizing broccoli harvest times, emphasizing the potential of RNN techniques in agricultural decision making.

Impact of deposition temperature on electrical properties of HZO-based FeRAM

This study presents a comprehensive investigation of the impact of the deposition temperature on the HfxZr1−xO2 (HZO) ferroelectric layer of ferroelectric random access memory with TaN electrodes. This investigation mainly focuses on its electrical characteristics and compares the differences. It is revealed that the deposition temperature plays a crucial role in determining the crystal structure of HZO, which can exhibit a combination of tetragonal and orthorhombic phases or exist solely in one of the two phases. Furthermore, the grain size of HZO varies with the deposition temperature. These findings correspond well to the electrical measurement results, including leakage current, polarization, capacitance, and reliability tests. The study tracks the phase transition process during the operation of switching cycles when the phase transition process can be monitored as well. To better understand the observed differences, physical models that shed light on the underlying mechanisms affected by deposition temperature are proposed at the end of the article.

Enhancing Human Activity Recognition in Smart Homes with Self-Supervised Learning and Self-Attention.

Deep learning models have gained prominence in human activity recognition using ambient sensors, particularly for telemonitoring older adults' daily activities in real-world scenarios. However, collecting large volumes of annotated sensor data presents a formidable challenge, given the time-consuming and costly nature of traditional manual annotation methods, especially for extensive projects. In response to this challenge, we propose a novel AttCLHAR model rooted in the self-supervised learning framework SimCLR and augmented with a self-attention mechanism. This model is designed for human activity recognition utilizing ambient sensor data, tailored explicitly for scenarios with limited or no annotations. AttCLHAR encompasses unsupervised pre-training and fine-tuning phases, sharing a common encoder module with two convolutional layers and a long short-term memory (LSTM) layer. The output is further connected to a self-attention layer, allowing the model to selectively focus on different input sequence segments. The incorporation of sharpness-aware minimization (SAM) aims to enhance model generalization by penalizing loss sharpness. The pre-training phase focuses on learning representative features from abundant unlabeled data, capturing both spatial and temporal dependencies in the sensor data. It facilitates the extraction of informative features for subsequent fine-tuning tasks. We extensively evaluated the AttCLHAR model using three CASAS smart home datasets (Aruba-1, Aruba-2, and Milan). We compared its performance against the SimCLR framework, SimCLR with SAM, and SimCLR with the self-attention layer. The experimental results demonstrate the superior performance of our approach, especially in semi-supervised and transfer learning scenarios. It outperforms existing models, marking a significant advancement in using self-supervised learning to extract valuable insights from unlabeled ambient sensor data in real-world environments.