Data-driven Architecture Research Articles

Data architectures and the prevalent cyberattacks encountered by West African higher educational institutions in the COVID-19 era

The growing interconnectedness of the world has led to a rise in cybersecurity risks. Although it is increasingly conventional to use technology to assist business transactions, exposure to these risks must be minimised to allow business owners to do transactions in a secure manner. While a wide range of studies have been undertaken regarding the effects of cyberattacks on several industries and sectors, However, very few studies have focused on the effects of cyberattacks on the educational sector, specifically higher educational institutions (HEIs) in West Africa. Consequently, this study developed a survey and distributed it to HEIs particularly universities in West Africa to examine the data architectures they employed, the cyberattacks they encountered during the COVID-19 pandemic period, and the role of data analysis in decision-making, as well as the countermeasures employed in identifying and preventing cyberattacks. A total of one thousand, one hundred and sixty-four (1164) responses were received from ninety-three (93) HEIs and analysed. According to the study’s findings, data-informed architecture was adopted by 71.8% of HEIs, data-driven architecture by 24.1%, and data-centric architecture by 4.1%, all of which were vulnerable to cyberattacks. In addition, there are further concerns around data analysis techniques, staff training gaps, and countermeasures for cyberattacks. The study’s conclusion includes suggestions for future research topics and recommendations for repelling cyberattacks in HEIs.

A probabilistic neural twin for treatment planning in peripheral pulmonary artery stenosis.

The substantial computational cost of high-fidelity models in numerical hemodynamics has, so far, relegated their use mainly to offline treatment planning. New breakthroughs in data-driven architectures and optimization techniques for fast surrogate modeling provide an exciting opportunity to overcome these limitations, enabling the use of such technology for time-critical decisions. We discuss an application to the repair of multiple stenosis in peripheral pulmonary artery disease through either transcatheter pulmonary artery rehabilitation or surgery, where it is of interest to achieve desired pressures and flows at specific locations in the pulmonary artery tree, while minimizing the risk for the patient. Since different degrees of success can be achieved in practice during treatment, we formulate the problem in probability, and solve it through a sample-based approach. We propose a new offline-online pipeline for probabilistic real-time treatment planning which combines offline assimilation of boundary conditions, model reduction, and training dataset generation with online estimation of marginal probabilities, possibly conditioned on the degree of augmentation observed in already repaired lesions. Moreover, we propose a new approach for the parametrization of arbitrarily shaped vascular repairs through iterative corrections of a zero-dimensional approximant. We demonstrate this pipeline for a diseased model of the pulmonary artery tree available through the Vascular Model Repository.

InVAErt networks: A data-driven framework for model synthesis and identifiability analysis

Use of generative models and deep learning for physics-based systems is currently dominated by the task of emulation. However, the remarkable flexibility offered by data-driven architectures would suggest to extend this representation to other aspects of system analysis including model inversion and identifiability. We introduce InVAErt (pronounced invert) networks, a comprehensive framework for data-driven analysis and synthesis of parametric physical systems which uses a deterministic encoder and decoder to represent the forward and inverse solution maps, a normalizing flow to capture the probabilistic distribution of system outputs, and a variational encoder designed to learn a compact latent representation for the lack of bijectivity between inputs and outputs. We formally analyze how changes in the penalty coefficients affect the stationarity condition of the loss function, the phenomenon of posterior collapse, and propose strategies for latent space sampling, since we find that all these aspects significantly affect both training and testing performance. We verify our framework through extensive numerical examples, including simple linear, nonlinear, and periodic maps, dynamical systems, and spatio-temporal PDEs.

Data-Driven Architecture to Encode Information in the Kinematics of Robots and Artificial Avatars

Data-Driven Architecture to Encode Information in the Kinematics of Robots and Artificial Avatars

Investigating permafrost carbon dynamics in Alaska with artificial intelligence

Positive feedbacks between permafrost degradation and the release of soil carbon into the atmosphere impact land–atmosphere interactions, disrupt the global carbon cycle, and accelerate climate change. The widespread distribution of thawing permafrost is causing a cascade of geophysical and biochemical disturbances with global impacts. Currently, few earth system models account for permafrost carbon feedback (PCF) mechanisms. This research study integrates artificial intelligence (AI) tools and information derived from field-scale surveys across the tundra and boreal landscapes in Alaska. We identify and interpret the permafrost carbon cycling links and feedback sensitivities with GeoCryoAI, a hybridized multimodal deep learning (DL) architecture of stacked convolutionally layered, memory-encoded recurrent neural networks (NN). This framework integrates in-situ measurements and flux tower observations for teacher forcing and model training. Preliminary experiments to quantify, validate, and forecast permafrost degradation and carbon efflux across Alaska demonstrate the fidelity of this data-driven architecture. More specifically, GeoCryoAI logs the ecological memory and effectively learns covariate dynamics while demonstrating an aptitude to simulate and forecast PCF dynamics—active layer thickness (ALT), carbon dioxide flux (CO2), and methane flux (CH4)—with high precision and minimal loss (i.e. ALTRMSE: 1.327 cm [1969–2022]; CO2 RMSE: 0.697 µmolCO2m−2s−1 [2003–2021]; CH4 RMSE: 0.715 nmolCH4m−2s−1 [2011–2022]). ALT variability is a sensitive harbinger of change, a unique signal characterizing the PCF, and our model is the first characterization of these dynamics across space and time.

Towards a Unified Pandemic Management Architecture: Survey, Challenges, and Future Directions

The pandemic caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) has impacted the economy, health, and society. Emerging strains are making pandemic management challenging. There is an urge to collect epidemiological, clinical, and physiological data to make an informed decision on mitigation. Advances in the Internet of Things (IoT) and edge computing provide solutions for pandemic management through data collection and intelligent computation. While existing data-driven architectures operate on specific application domains and attempt to automate decision-making, they do not capture the multifaceted interaction among computational models, communication infrastructure, and data. In this article, we survey the existing approaches for pandemic management, including data repositories and contact-tracing applications. We envision a unified pandemic management architecture that leverages the IoT and edge computing paradigms to automate recommendations on vaccine distribution, dynamic lockdown, mobility scheduling, and pandemic trend prediction. We elucidate the data flow among the layers, namely, cloud, edge, and end device layers. Moreover, we address the privacy implications, threats, regulations, and solutions that may be adapted to optimize the utility of health data with security guarantees. The article ends with a discussion of the limitations of the architecture and research directions to enhance its practicality.

Standardizing chemical compounds with language models

With the growing amount of chemical data stored digitally, it has become crucial to represent chemical compounds accurately and consistently. Harmonized representations facilitate the extraction of insightful information from datasets, and are advantageous for machine learning applications. To achieve consistent representations throughout datasets, one relies on molecule standardization, which is typically accomplished using rule-based algorithms that modify descriptions of functional groups. Here, we present the first deep-learning model for molecular standardization. We enable custom standardization schemes based solely on data, which, as additional benefit, support standardization options that are difficult to encode into rules. Our model achieves over accuracy in learning two popular rule-based standardization protocols. We then follow a transfer learning approach to standardize metal-organic compounds (for which there is currently no automated standardization practice), based on a human-curated dataset of 1512 compounds. This model predicts the expected standardized molecular format with a test accuracy of 80.7%. As standardization can be considered, more broadly, a transformation from undesired to desired representations of compounds, the same data-driven architecture can be applied to other tasks. For instance, we demonstrate the application to compound canonicalization and to the determination of major tautomers in solution, based on computed and experimental data.

What issues are data scientists talking about? Identification of current data science issues using semantic content analysis of Q&A communities.

Because of the growing involvement of communities from various disciplines, data science is constantly evolving and gaining popularity. The growing interest in data science-based services and applications presents numerous challenges for their development. Therefore, data scientists frequently turn to various forums, particularly domain-specific Q&A websites, to solve difficulties. These websites evolve into data science knowledge repositories over time. Analysis of such repositories can provide valuable insights into the applications, topics, trends, and challenges of data science. In this article, we investigated what data scientists are asking by analyzing all posts to date on DSSE, a data science-focused Q&A website. To discover main topics embedded in data science discussions, we used latent Dirichlet allocation (LDA), a probabilistic approach for topic modeling. As a result of this analysis, 18 main topics were identified that demonstrate the current interests and issues in data science. We then examined the topics' popularity and difficulty. In addition, we identified the most commonly used tasks, techniques, and tools in data science. As a result, "Model Training", "Machine Learning", and "Neural Networks" emerged as the most prominent topics. Also, "Data Manipulation", "Coding Errors", and "Tools" were identified as the most viewed (most popular) topics. On the other hand, the most difficult topics were identified as "Time Series", "Computer Vision", and "Recommendation Systems". Our findings have significant implications for many data science stakeholders who are striving to advance data-driven architectures, concepts, tools, and techniques.

Frequency learning attention networks based on deep learning for automatic modulation classification in wireless communication

Deep neural networks have been recently applied in automatic modulation classification task and achieved remarkable success. However, Existing neural networks mainly focus on the purely data-driven architecture design, and fail to explore the hand-crafted feature mechanisms which are particularly significant for radio signal presentation in wireless communication. Inspired by digital signal processing theories, we propose frequency learning attention networks (FLANs) to analyze the radio spectral bias from frequency perspective, based on a multi-spectral attention mechanism for learning-based frequency components selection. FLANs are the general case of classical global average pooling and leverage identical structures of the popular neural networks. Extensive experiments have been conducted to validate the superiority of FLANs for automatic modulation classification over a wide variety of state-of-the-art methods on RADIOML 2018.01A dataset.

Long Short-Term Memory-Based Neural Networks for Missile Maneuvers Trajectories Prediction⋆

Due to its extensive applications in different contexts, moving target tracking has become a hot topic in the last years, above all in the military field. Specifically, missile tracking research received a great effort, mainly for its importance in terms of security and safety. Herein, traditional solutions, e.g. Interacting Multiple Model (IMM) based on the Kalman estimation theory, achieve good performance under the main restrictive assumption of the a priori knowledge of the target model, so neglecting the unavoidable presence of model uncertainties and limiting the achievable tracking accuracy only by the presence of the measurement noise. With the specific aim of overcoming this narrowness, this work investigates the capability of deep neural networks in predicting the missile maneuvering trajectories in a model-free fashion. The idea is to leverage the Long-Short Term Memory (LSTM) net due to its excellent capability in learning long-term dependencies of temporal information. Two different LSTM-based architectures have been hence designed to predict both position and velocity of a missile using raw and noisy measurements provided by a realistic radar system, exploiting a large database abundant of realistic off-line data. Training results and theoretical derivations are verified through non-trivial scenarios in order to assess the capability of predicting unknown and realistic 3D missile maneuvers. Finally, the proposed approach has been also compared with a performing model-based IMM algorithm, suitably tuned to deal with realistic missile maneuvers, confirming the excellent generalization abilities of the developed data-driven architectures for different datasets.

Towards socially-competent and culturally-adaptive artificial agents

AbstractThe development of artificial agents for social interaction pushes to enrich robots with social skills and knowledge about (local) social norms. One possibility is to distinguish the expressive and the functional orders during a human-robot interaction. The overarching aim of this work is to set a framework to make the artificial agent socially-competent beyond dyadic interaction – interaction in varying multi-party social situations – and beyond individual-based user personalization, thereby enlarging the current conception of “culturally-adaptive”. The core idea is to provide the artificial agent with the capability to handle different kinds of interactional disruptions, and associated recovery strategies, in microsociology. The result is obtained by classifying functional and social disruptions, and by investigating the requirements a robot’s architecture should satisfy to exploit such knowledge. The paper also highlights how this level of competence is achieved by focusing on just three dimensions: (i) social capability, (ii) relational role, and (iii) proximity, leaving aside the further complexity of full-fledged human-human interactions. Without going into technical aspects, End-to-end Data-driven Architectures and Modular Architectures are discussed to evaluate the degree to which they can exploit this new set of social and cultural knowledge. Finally, a list of general requirements for such agents is proposed.

Data-Driven Intelligent Manipulation of Particles in Microfluidics.

Automated manipulation of small particles using external (e.g., magnetic, electric and acoustic) fields has been an emerging technique widely used in different areas. The manipulation typically necessitates a reduced-order physical model characterizing the field-driven motion of particles in a complex environment. Such models are available only for highly idealized settings but are absent for a general scenario of particle manipulation typically involving complex nonlinear processes, which has limited its application. In this work, the authorspresent a data-driven architecture for controlling particles in microfluidics based on hydrodynamic manipulation. The architecture replaces the difficult-to-derive model by a generally trainable artificial neural network to describe the kinematics of particles, and subsequently identifies the optimal operations to manipulate particles. The authorssuccessfully demonstrate a diverse set of particle manipulations in a numerically emulated microfluidic chamber, including targeted assembly of particles and subsequent navigation of the assembled cluster, simultaneous path planning for multiple particles, and steering one particle through obstacles. Theapproach achieves both spatial and temporal controllability of high precision for these settings. This achievement revolutionizes automated particle manipulation, showing the potential of data-driven approaches and machine learning in improving microfluidic technologies for enhanced flexibility andintelligence.

Machine-learning-based data-driven discovery of nonlinear phase-field dynamics.

One of the main questions regarding complex systems at large scales concerns the effective interactions and driving forces that emerge from the detailed microscopic properties. Coarse-grained models aim to describe complex systems in terms of coarse-scale equationswith a reduced number of degrees of freedom. Recent developments in machine-learning algorithms have significantly empowered the discovery process of governing equationsdirectly from data. However, it remains difficult to discover partial differential equations(PDEs) with high-order derivatives. In this paper, we present data-driven architectures based on a multilayer perceptron, a convolutional neural network (CNN), and a combination of a CNN and long short-term memory structures for discovering the nonlinear equationsof motion for phase-field models with nonconserved and conserved order parameters. The well-known Allen-Cahn, Cahn-Hilliard, and phase-field crystal models were used as test cases. Two conceptually different types of implementations were used: (a) guided by physical intuition (such as the local dependence of the derivatives) and (b) in the absence of any physical assumptions (black-box model). We show that not only can we effectively learn the time derivatives of the field in both scenarios, but we can also use the data-driven PDEs to propagate the field in time and achieve results in good agreement with the original PDEs.

Understanding user concerns and choice of app architectures in designing Audio-based mHealth Apps

Due to recent advancements in smartphone sensing and computing capabilities, artificially intelligent mobile health (mHealth) applications are getting popular to monitor a range of diseases, including coronavirus-caused COVID-19 disease, chronic obstructive pulmonary disease (COPD), asthma, bronchitis, emphysema, and sleep apnea, among many others utilizing the audio recordings obtained from the smartphone microphones. Compared to other mHealth apps, audio-based apps suffer from various user concerns, including the privacy of data and battery drain rate , among many other concerns, which can adversely affect the user compliance, and app utility and life cycle. To address user concerns, mHealth apps should provide users options to configure the app, i.e., choose an architecture from a set of options based on user concerns and preferences. However, there is a dearth of knowledge about audio-based health monitoring app design. In this work, we present a focused user-centric app design study to better understand various concerns and choices that users want to see in an audio-based mHealth app. From a detailed analysis of 60 subjects with varying backgrounds, we find that around 85% subjects are concerned about the privacy of data and 93% subjects prefer to pick an app architecture that will not send raw audio recordings to a server. Findings from this work can guide the design of future mHealth apps that utilizes privacy-sensitive audio data. • Determine the major user concerns about the audio-based health monitoring smartphone apps and user choice when provided with a set of app configurations, i.e., architectures, that a user can choose when using a health monitoring app that relies on continuous microphone sensing to obtain audio recordings. • People with diferent backgrounds have a similar level of concerns while reporting their concerns about audio-based health monitoring apps. • Most of the study paracipants (around 66%) choose either the complete on-phone architecture, i.e., architecture that performs all computations on the phone, or the non- speech data-driven architecture, i.e., architecture that removes all speech components from audio recordings in the phone before sending them to the server for additional computation. • Only 7% participants choose the raw data-driven architecture, i.e., architecture that sends raw audio recordings to the server for all computation. • These fndings could be used as guidelines for the future user-centric app design to address user concerns and provide app customization options with the goal to improve the utility and compliance of the mHealth apps that rely on audio data obtained from the continous microphone sensing.

NCare: Fault-aware edge intelligence for rendering viable sensor nodes

In remote sensor nodes, energy harvesters are used with rechargeable batteries as the power source. As these batteries undergo capacitance reduction, they degrade with each charge–discharge cycle. Most of the existing researches do not consider edge-based computation for battery health prediction. This paper proposes a smart, accurate, and reliable data-driven architecture, nCare, to predict the battery’s health at low power and low computation costs. Three models are used for health prediction, two Gaussian Process Regression (GPR) models deployed at the sensor node — one to predict the State of Health (SoH) of the battery for the first few cycles and the second to predict the Remaining Useful Life (RUL) of battery. An AutoKeras model deployed at the gateway where it will be trained using the first 200 cycles. Then weight matrix and layer information will be sent to sensor nodes to replicate the model without training to predict SoH after 200 cycles. Two main parameters, battery temperature and elapsed time (ET), are used to predict the SoH, which is further used for RUL prediction. Furthermore, to prove the efficiency of the RUL prediction method, the calculated Absolute Error (AE), Relative Mean Square Error (RMSE), and Relative error (RE) are 0.019, 0.142, and 0.005, respectively for a battery of 730 cycles. This method helps the sensor node acquire intelligence that will be battery specific. Blockchain is used for maintaining the data integrity of the architecture. nCare thus ensures an effective and accurate solution for predicting the RUL of a battery.

Development of a hybrid single-photon imaging detector with embedded CMOS pixelated anode

The development of a single-photon detector based on a vacuum tube, transmission photocathode, microchannel plate and CMOS pixelated read-out anode is presented. This imager will be capable of detecting up to 1 billion photons per second over an area of 7 cm2, with simultaneous measurement of position and time with resolutions of about 5 microns and few tens of picosecond, respectively. The detector has embedded pulse-processing electronics with data-driven architecture, based on the Timepix4 ASIC, producing up to 160 Gb/s data that will be handled by a high-throughput FPGA-based external electronics and data acquisition system. These performances will enable significant advances in particle physics, life sciences, quantum optics or other emerging fields where the detection of single photons with excellent timing and position resolutions are simultaneously required.

A Data-Driven Approach for University Public Opinion Analysis and Its Applications

In the era of mobile Internet, college students increasingly tend to express their opinions and views through online social media; furthermore, social media influence the value judgments of college students. Therefore, it is vital to understand and analyze university online public opinion over time. In this paper, we propose a data-driven architecture for analysis of university online public opinion. Weibo, WeChat, Douyin, Zhihu and Toutiao apps are selected as sources for collection of public opinion data. Crawler technology is utilized to automatically obtain user data about target topics to form a database. To avoid the drawbacks of traditional methods, such as sentiment lexicon and machine learning, which rely on a priori knowledge and complex handcrafted features, the Word2Vec tool is used to perform word embedding, the LSTM-CFR model is proposed to realize Chinese word segmentation and a convolutional neural network (CNN) is built to automatically extract implicit features in word vectors, ultimately establishing the nonlinear relationships between implicit features and the sentiment tendency of university public opinion. The experimental results show that the proposed model is more accurate than SVM, RF, NBC and GMM methods, providing valuable information with respect to public opinion management.

Grid adaptive reduced-order model of fluid flow based on graph convolutional neural network

In the interdisciplinary field of data-driven models and computational fluid mechanics, the reduced-order model for flow field prediction is mainly constructed by a convolutional neural network (CNN) in recent years. However, the standard CNN is only applicable to data with Euclidean spatial structure, while data with non-Euclidean properties can only be convolved after pixelization, which usually leads to decreased data accuracy. In this work, a novel data-driven framework based on graph convolution network (GCN) is proposed to allow the convolution operator to predict fluid dynamics on non-uniform structured or unstructured mesh data. This is achieved by the fact that the graph data inherit the spatial characteristics of the mesh and by the message passing mechanism of GCN. The conversion method from the form of mesh data to graph data and the operation mechanism of GCN are clarified. Moreover, additional relevance features and weight loss function of the dataset are also investigated to improve the model performance. The model learns an end-to-end mapping between the mesh spatial features and the physical flow field. Through our studies of various cases of internal flow, it is shown that the proposed GCN-based model offers excellent adaptability to non-uniformly distributed mesh data, while also achieving a high accuracy and three-order speedup compared with numerical simulation. Our framework generalizes the graph convolution network to flow field prediction and opens the door to further extending GCN to most existing data-driven architectures of fluid dynamics in the future.

Piano Automatic Composition and Quantitative Perception under the Data-Driven Architecture

This paper combines automatic piano composition with quantitative perception, extracts note features from the demonstration audio, and builds a neural network model to complete automatic composition. First of all, in view of the diversity and complexity of the data collected in the quantitative perception of piano automatic composition, the energy efficiency-related state data of the piano automatic composition operation is collected, carried out, and dealt with. Secondly, a perceptual data-driven energy efficient evaluation and decision-making method is proposed. This method is based on time series index data. After determining the time subjective weight through time entropy, the time dimension factor is introduced, and then the subjective time weight is adjusted by the minimum variance method. Then, we consider the impact of the perception period on the perception efficiency and accuracy, calculate and dynamically adjust the perception period based on the running data, consider the needs of the perception object in different scenarios, and update the perception object in real time during the operation. Finally, combined with the level weights determined by the data-driven architecture, the dynamic manufacturing capability index and energy efficiency index of the equipment are finally obtained. The energy efficiency evaluation of the manufacturing system of the data-driven architecture proves the feasibility and scientificity of the evaluation method and achieves the goal of it. The simulation experiment results show that it can reduce the perception overhead while ensuring the perception efficiency and accuracy.

Aircraft engine remaining useful life estimation via a double attention-based data-driven architecture

Remaining useful life (RUL) estimation has been intensively studied, given its important role in prognostics and health management (PHM) of industry. Recently, data-driven structures such as convolutional neural networks (CNNs), have achieved outstanding RUL prediction performance. However, conventional CNNs do not include an adequate mechanism for adaptively weighing input features. In this paper, we propose a double attention-based data-driven framework for aircraft engine RUL prognostics. Specifically, a channel attention-based CNN was utilized to apply greater weights to more significant features. Next, a Transformer was used to focus attention on these features at critical time steps. We validated the effectiveness of the proposed framework on benchmark datasets for aircraft engine RUL estimation. The experimental results indicate that the proposed double attention-based architecture outperformed the existing state-of-the-art (SOTA) algorithms. The double attention-based RUL prediction method can detect the risk of equipment failure and reduce loss.