UMBC Center for Accelerated Real Time Analysis

Permanent URI for this collectionhttp://hdl.handle.net/11603/26415

Real time analytics is the leading edge of a smart data revolution, pushed by Internet advances in sensor hardware on one side and AI/ML streaming acceleration on the other. Center for Accelerated Real Time Analytics (CARTA) explores the realm streaming applications of Magna Analytics. The center works with next-generation hardware technologies, like the IBM Minsky with onboard GPU accelerated processors and Flash RAM, a Smart Cyber Physical Sensor Systems to build Cognitive Analytics systems and Active storage devices for real time analytics. This will lead to the automated ingestion and simultaneous analytics of Big Datasets generated in various domains including Cyberspace, Healthcare, Internet of Things (IoT) and the Scientific arena, and the creation of self learning, self correcting “smart” systems.

Browse

Now showing 1 - 20 of 67

A Framework for Empirical Fourier Decomposition based Gesture Classification for Stroke Rehabilitation
(IEEE, 2024-11-11) Chen, Ke; Wang, Honggang; Catlin, Andrew; Satyanarayana, Ashwin; Vinjamuri, Ramana; Kadiyala, Sai Praveen
The demand for surface electromyography (sEMG) based exoskeletons is rapidly increasing due to their non-invasive nature and ease of use. With increase in use of Internet-of-Things (IoT) based devices in daily life, there is a greater acceptance of exoskeleton based rehab. As a result, there is a need for highly accurate and generalizable gesture classification mechanisms based on sEMG data. In this work, we present a framework which pre-processes raw sEMG signals with Empirical Fourier Decomposition (EFD) based approach followed by dimension reduction. This resulted in improved performance of the hand gesture classification. EFD decomposition’s efficacy of handling mode mixing problem on non-stationary signals, resulted in less number of decomposed components. In the next step, a thorough analysis of decomposed components as well as inter-channel analysis is performed to identify the key components and channels that contribute towards the improved gesture classification accuracy. As a third step, we conducted ablation studies on time-domain features to observe the variations in accuracy on different models. Finally, we present a case study of comparison of automated feature extraction based gesture classification vs. manual feature extraction based methods. Experimental results show that manual feature based gesture classification method thoroughly outperformed automated feature extraction based methods, thus emphasizing a need for rigorous fine tuning of automated models.
Tutorial on Causal Inference with Spatiotemporal Data
(ACM, 2024-11-04) Ali, Sahara; Wang, Jianwu
Spatiotemporal data, which captures how variables evolve across space and time, is ubiquitous in fields such as environmental science, epidemiology, and urban planning. However, identifying causal relationships in these datasets is challenging due to the presence of spatial dependencies, temporal autocorrelation, and confounding factors. This tutorial provides a comprehensive introduction to spatiotemporal causal inference, offering both theoretical foundations and practical guidance for researchers and practitioners. We explore key concepts such as causal inference frameworks, the impact of confounding in spatiotemporal settings, and the challenges posed by spatial and temporal dependencies. The paper covers synthetic spatiotemporal benchmark data generation, widely used spatiotemporal causal inference techniques, including regression-based, propensity score-based, and deep learning-based methods, and demonstrates their application using synthetic datasets. Through step-by-step examples, readers will gain a clear understanding of how to address common challenges and apply causal inference techniques to spatiotemporal data. This tutorial serves as a valuable resource for those looking to improve the rigor and reliability of their causal analyses in spatiotemporal contexts.
Accelerating Subglacial Bed Topography Prediction in Greenland: A Performance Evaluation of Spark-Optimized Machine Learning Models
(2024) Cham, Mostafa; Tabassum, Tartela; Shakeri, Ehsan; Wang, Jianwu
Identifying neurophysiological correlates of stress
(frontiers, 2024-10-24) Pei, Dingyi; Tirumala, Shravika; Tun, Kyaw T.; Ajendla, Akshara; Vinjamuri, Ramana
Stress has been recognized as a pivotal indicator which can lead to severe mental disorders. Persistent exposure to stress will increase the risk for various physical and mental health problems. Early and reliable detection of stress-related status is critical for promoting wellbeing and developing effective interventions. This study attempted multi-type and multi-level stress detection by fusing features extracted from multiple physiological signals including electroencephalography (EEG) and peripheral physiological signals. Eleven healthy individuals participated in validated stress-inducing protocols designed to induce social and mental stress and discriminant multi-level and multi-type stress. A range of machine learning methods were applied and evaluated on physiological signals of various durations. An average accuracy of 98.1% and 97.8% was achieved in identifying stress type and stress level respectively, using 4-s neurophysiological signals. These findings have promising implications for enhancing the precision and practicality of real-time stress monitoring applications.
Investigating Causal Cues: Strengthening Spoofed Audio Detection with Human-Discernible Linguistic Features
(2024-09-09) Khanjani, Zahra; Ale, Tolulope; Wang, Jianwu; Davis, Lavon; Mallinson, Christine; Janeja, Vandana
Several types of spoofed audio, such as mimicry, replay attacks, and deepfakes, have created societal challenges to information integrity. Recently, researchers have worked with sociolinguistics experts to label spoofed audio samples with Expert Defined Linguistic Features (EDLFs) that can be discerned by the human ear: pitch, pause, word-initial and word-final release bursts of consonant stops, audible intake or outtake of breath, and overall audio quality. It is established that there is an improvement in several deepfake detection algorithms when they augmented the traditional and common features of audio data with these EDLFs. In this paper, using a hybrid dataset comprised of multiple types of spoofed audio augmented with sociolinguistic annotations, we investigate causal discovery and inferences between the discernible linguistic features and the label in the audio clips, comparing the findings of the causal models with the expert ground truth validation labeling process. Our findings suggest that the causal models indicate the utility of incorporating linguistic features to help discern spoofed audio, as well as the overall need and opportunity to incorporate human knowledge into models and techniques for strengthening AI models. The causal discovery and inference can be used as a foundation of training humans to discern spoofed audio as well as automating EDLFs labeling for the purpose of performance improvement of the common AI-based spoofed audio detectors.
Atmospheric Gravity Wave Detection Using Transfer Learning Techniques
(IEEE, 2022-12) González, Jorge López; Chapman, Theodore; Chen, Kathryn; Nguyen, Hannah; Chambers, Logan; Mostafa, Seraj Al Mahmud; Wang, Jianwu; Purushotham, Sanjay; Wang, Chenxi; Yue, Jia
Atmospheric gravity waves are produced when gravity attempts to restore disturbances through stable layers in the atmosphere. They have a visible effect on many atmospheric phenomena such as global circulation and air turbulence. Despite their importance, however, little research has been conducted on how to detect gravity waves using machine learning algorithms. We faced two major challenges in our research: our raw data had a lot of noise and the labeled dataset was extremely small. In this study, we explored various methods of preprocessing and transfer learning in order to address those challenges. We pre-trained an autoencoder on unlabeled data before training it to classify labeled data. We also created a custom CNN by combining certain pre-trained layers from the InceptionV3 Model trained on ImageNet with custom layers and a custom learning rate scheduler. Experiments show that our best model outperformed the best performing baseline model by 6.36% in terms of test accuracy.
Integrating Electroencephalography Source Localization and Residual Convolutional Neural Network for Advanced Stroke Rehabilitation
(MDPI, 2024-09-27) Kaviri, Sina Makhdoomi; Vinjamuri, Ramana
Motor impairments caused by stroke significantly affect daily activities and reduce quality of life, highlighting the need for effective rehabilitation strategies. This study presents a novel approach to classifying motor tasks using EEG data from acute stroke patients, focusing on left-hand motor imagery, right-hand motor imagery, and rest states. By using advanced source localization techniques, such as Minimum Norm Estimation (MNE), dipole fitting, and beamforming, integrated with a customized Residual Convolutional Neural Network (ResNetCNN) architecture, we achieved superior spatial pattern recognition in EEG data. Our approach yielded classification accuracies of 91.03% with dipole fitting, 89.07% with MNE, and 87.17% with beamforming, markedly surpassing the 55.57% to 72.21% range of traditional sensor domain methods. These results highlight the efficacy of transitioning from sensor to source domain in capturing precise brain activity. The enhanced accuracy and reliability of our method hold significant potential for advancing brain–computer interfaces (BCIs) in neurorehabilitation. This study emphasizes the importance of using advanced EEG classification techniques to provide clinicians with precise tools for developing individualized therapy plans, potentially leading to substantial improvements in motor function recovery and overall patient outcomes. Future work will focus on integrating these techniques into practical BCI systems and assessing their long-term impact on stroke rehabilitation.
Integrating Fourier Transform and Residual Learning for Arctic Sea Ice Forecasting
(IEEE, 2023-12) Lapp, Louis; Ali, Sahara; Wang, Jianwu
Arctic sea ice plays integral roles in both polar and global environmental systems, notably ecosystems, commu-nities, and economies. As sea ice continues to decline due to climate change, it has become imperative to accurately predict the future of sea ice extent (SIE). Using datasets of Arctic meteorological and SIE variables spanning 1979 to 2021, we propose architectures capable of processing multivariate time series and spatiotemporal data. Our proposed framework consists of ensembled stacked Fourier Transform signals (FFTstack) and Gradient Boosting models. In FFTstack, grid search iteratively detects the optimal combination of representative FFT signals, a process that improves upon current FFT implementations and deseasonalizers. An optimized Gradient Boosting Regressor is then trained on the residual of the FFTstack output. Through ex-periment, we found that the models trained on both multivariate and spatiotemporal time series data performed either similar to or better than models in existing research. In addition, we found that integration of FFTstack improves the performance of current multivariate time series deep learning models. We conclude that the high flexibility and performance of this methodology have promising applications in guiding future adaptation, resilience, and mitigation efforts in response to Arctic sea ice retreat.
Hybrid Ensemble Deep Graph Temporal Clustering for Spatiotemporal Data
(2024-09-19) Nji, Francis Ndikum; Faruque, Omar; Cham, Mostafa; Janeja, Vandana; Wang, Jianwu
Classifying subsets based on spatial and temporal features is crucial to the analysis of spatiotemporal data given the inherent spatial and temporal variability. Since no single clustering algorithm ensures optimal results, researchers have increasingly explored the effectiveness of ensemble approaches. Ensemble clustering has attracted much attention due to increased diversity, better generalization, and overall improved clustering performance. While ensemble clustering may yield promising results on simple datasets, it has not been fully explored on complex multivariate spatiotemporal data. For our contribution to this field, we propose a novel hybrid ensemble deep graph temporal clustering (HEDGTC) method for multivariate spatiotemporal data. HEDGTC integrates homogeneous and heterogeneous ensemble methods and adopts a dual consensus approach to address noise and misclassification from traditional clustering. It further applies a graph attention autoencoder network to improve clustering performance and stability. When evaluated on three real-world multivariate spatiotemporal data, HEDGTC outperforms state-of-the-art ensemble clustering models by showing improved performance and stability with consistent results. This indicates that HEDGTC can effectively capture implicit temporal patterns in complex spatiotemporal data.
Flood-ResNet50: Optimized Deep Learning Model for Efficient Flood Detection on Edge Device
(IEEE, 2024-03-19) Khan, Md Azim; Ahmed, Nadeem; Padela, Joyce; Raza, Muhammad Shehrose; Gangopadhyay, Aryya; Wang, Jianwu; Foulds, James; Busart, Carl; Erbacher, Robert F.
Floods are highly destructive natural disasters that result in significant economic losses and endanger human and wildlife lives. Efficiently monitoring Flooded areas through the utilization of deep learning models can contribute to mitigating these risks. This study focuses on the deployment of deep learning models specifically designed for classifying flooded and non-flooded in UAV images. In consideration of computational costs, we propose modified version of ResNet50 called Flood-ResNet50. By incorporating additional layers and leveraging transfer learning techniques, Flood-ResNet50 achieves comparable performance to larger models like VGG16/19, AlexNet, DenseNet161, EfficientNetB7, Swin(small), and vision transformer. Experimental results demonstrate that the proposed modification of ResNet50, incorporating additional layers, achieves a classification accuracy of 96.43%, F1 score of 86.36%, Recall of 81.11%, Precision of 92.41 %, model size 98MB and FLOPs 4.3 billions for the FloodNet dataset. When deployed on edge devices such as the Jetson Nano, our model demonstrates faster inference speed (820 ms), higher throughput (39.02 fps), and lower average power consumption (6.9 W) compared to larger ResNet101 and ResNet152 models.
Variability of Eastern North Atlantic Summertime Marine Boundary Layer Clouds and Aerosols Across Different Synoptic Regimes Identified with Multiple Conditions
(2024-08-22) Zheng, Xue; Qiu, Shaoyue; Zhang, Damao; Adebiyi, Adeyemi A.; Zheng, Xiaojian; Faruque, Omar; Tao, Cheng; Wang, Jianwu
This study estimates the meteorological covariations of aerosol and marine boundary layer (MBL) cloud properties in the Eastern North Atlantic (ENA) region, characterized by diverse synoptic conditions. Using a deep-learning-based clustering model with mid-level and surface daily meteorological data, we identify seven distinct synoptic regimes during the summer from 2016 to 2021. Our analysis, incorporating reanalysis data and satellite retrievals, shows that surface aerosols and MBL clouds exhibit clear regime-dependent characteristics, while lower tropospheric aerosols do not. This discrepancy likely arises synoptic regimes determined by daily large-scale conditions may overlook air mass histories that predominantly dictate lower tropospheric aerosol conditions. Focusing on three regimes dominated by northerly winds, we analyze the Atmospheric Radiation Measurement Program (ARM) ENA observations on Graciosa Island in the Azores. In the subtropical anticyclone regime, fewer cumulus clouds and more single-layer stratocumulus clouds with light drizzles are observed, along with the highest cloud droplet number concentration (Nd), surface Cloud Condensation Nuclei (CCN) and surface aerosol levels. The post-trough regime features more broken or multi-layer stratocumulus clouds with slightly higher surface rain rate, and lower Nd and surface CCN levels. The weak trough regime is characterized by the deepest MBL clouds, primarily cumulus and broken stratocumulus clouds, with the strongest surface rain rate and the lowest Nd, surface CCN and surface aerosol levels, indicating strong wet scavenging. These findings highlight the importance of considering the covariation of cloud and aerosol properties driven by large-scale regimes when assessing aerosol indirect effects using observations.
Generative AI Based Difficulty Level Design of Serious Games for Stroke Rehabilitation
(IEEE, 2024-08-28) Chen, Ke; Vinjamuri, Ramana; Wang, Honggang; Kadiyala, Sai Praveen
Internet-of-Things (IoT) based solutions are gaining momentum in delivering efficient solutions in health care domain, reducing financial and physical burden on patients and improving ease of treatment for physicians. One such smart health care solution are Serious games. Serious games aid rehabilitation in various fields. For physical rehabilitation, personalization is important for improving training results. A scientific approach for difficulty level design can facilitate players to get effective rehabilitation. The automation of personalized difficulty level design helps the self-guided game-based rehabilitation approach, become simplified and efficient. AI is advancing the design of personalized serious game for rehabilitation through data-driven and individual-oriented methods. In this work, we present Generative AI based design of gamified training plan, especially difficulty level plan which could go beyond rule based solutions. We apply Generative Adversarial Networks (GANs) to address the problem arising from large sequential data and variable requirement. This helps to overcome the limitation of unrealistic long term practice session for a rehabilitation patient by simplifying the training time. When compared with the results from Long Short-Term Memory (LSTM) based approach, our GANs based approach gave a 4.5X less variation in difficulty level and 6.5X less loss which proved the efficacy of our proposed approach in generating accurate difficulty levels. When compared with existing literature our proposed work simultaneously performs better on various parameters namely faster convergence, minimum emphasis on past performance of players, low data requirement for training and demographic flexibility.
gWaveNet: Classification of Gravity Waves from Noisy Satellite Data using Custom Kernel Integrated Deep Learning Method
(2024-08-26) Mostafa, Seraj Al Mahmud; Faruque, Omar; Wang, Chenxi; Yue, Jia; Purushotham, Sanjay; Wang, Jianwu
Atmospheric gravity waves occur in the Earths atmosphere caused by an interplay between gravity and buoyancy forces. These waves have profound impacts on various aspects of the atmosphere, including the patterns of precipitation, cloud formation, ozone distribution, aerosols, and pollutant dispersion. Therefore, understanding gravity waves is essential to comprehend and monitor changes in a wide range of atmospheric behaviors. Limited studies have been conducted to identify gravity waves from satellite data using machine learning techniques. Particularly, without applying noise removal techniques, it remains an underexplored area of research. This study presents a novel kernel design aimed at identifying gravity waves within satellite images. The proposed kernel is seamlessly integrated into a deep convolutional neural network, denoted as gWaveNet. Our proposed model exhibits impressive proficiency in detecting images containing gravity waves from noisy satellite data without any feature engineering. The empirical results show our model outperforms related approaches by achieving over 98% training accuracy and over 94% test accuracy which is known to be the best result for gravity waves detection up to the time of this work. We open sourced our code at https://rb.gy/qn68ku.
UMBC expands its next-generation tech analytics research with Phase II of the Center for Accelerated Real Time Analytics
(UMBC News, 2024-01-17) Fraser, Adriana
MASON: A Model for Adapting Service-Oriented Grid Applications
(Springer, 2003) Li, Gang; Wang, Jianwu; Wang, Jing; Han, Yanbo; Zhao, Zhuofeng; Wagner, Roland M.; Hu, Haitao
Service-oriented computing, which offers more flexible means for application development, is gaining popularity. Service-oriented grid applications are constructed by selecting and composing appropriate services. They are one kind of promising applications in grid environments. However, the dynamism and autonomy of environments make the issues of dynamically adapting a service-oriented grid application urgent. This paper brings forward a model that supports not only monitoring applications through gathering and managing state and structure metadata of service-oriented grid applications, but also dynamic application adjustment by changing the metadata. Besides that, the realization and application of the model is presented also.
An Approach to Abstracting and Transforming Web Services for End-user-doable Construction of Service-Oriented Applications
(Gesellschaft für Informatik, 2005) Yu, Jian; Fang, Jun; Han, Yanbo; Wang, Jianwu; Zhang, Cheng
End-user-programmable business-level services composition is an effective way to build virtual organizations of individual applications in a just-intime manner. Challenging issues include how to model business-level services so that the end users can understand and compose them; how to associate businesslevel services to underlying Web services. This paper presents a service virtualization approach called VINCA Virualization to supporting the abstraction, transformation, binding and execution of Web services by end users. Four key mechanisms of VINCA Virualization namely semantics annotation, services aggregation, virtualization operation and services convergence are discussed in details. VINCA Virualization has been implemented and its application in a real-world project is illustrated. The paper concludes with a comparative study with other related works.
An Approach to Domain-Specific Reuse in Service-Oriented Environments
(Springer, 2008) Wang, Jianwu; Yu, Jian; Falcarin, Paolo; Han, Yanbo; Morisio, Maurizio
Domain engineering is successful in promoting reuse. An approach to domain-specific reuse in service-oriented environments is proposed to facilitate service requesters to reuse Web services. In the approach, we present a conceptual model of domain-specific services (called domain service). Domain services in a certain business domain are modeled by semantic and feature modeling techniques, and bound to Web services with diverse capabilities through a variability-supported matching mechanism. By reusing pre-modeled domain services, service requesters can describe their requests easily through a service customization mechanism. Web service selection based on customized results can also be optimized by reusing the pre-matching results between domain services and Web services. Feasibility of the whole approach is demonstrated on an example.
A High-Level Distributed Execution Framework for Scientific Workflows
(2008) Wang, Jianwu; Altintas, Ilkay; Berkley, Chad; Gilbert, Lucas; Jones, Matthew B.
Accelerating Parameter Sweep Workflows by Utilizing Ad-hoc Network Computing Resources: an Ecological Example
Wang, Jianwu; Altintas, Ilkay; Hosseini, Parviez R.; Barseghian, Derik; Crawl, Daniel; Berkley, Chad; Jones, Matthew B.
Facilitating e-Science Discovery Using Scientific Workflows on the Grid
(Springer, 2011-01-01) Wang, Jianwu; Korambath, Prakashan; Kim, Seonah; Johnson, Scott; Jin, Kejian; Crawl, Daniel; Altintas, Ilkay; Smallen, Shava; Labate, Bill; Houk, Kendall N.
e-Science has been greatly enhanced from the developing capability and usability of cyberinfrastructure. This chapter explains how scientific workflow systems can facilitate e-Science discovery in Grid environments by providing features including scientific process automation, resource consolidation, parallelism, provenance tracking, fault tolerance, and workflow reuse. We first overview the core services to support e-Science discovery. To demonstrate how these services can be seamlessly assembled, an open source scientific workflow system, called Kepler, is integrated into the University of California Grid. This architecture is being applied to a computational enzyme design process, which is a formidable and collaborative problem in computational chemistry that challenges our knowledge of protein chemistry. Our implementation and experiments validate how the Kepler workflow system can make the scientific computation process automated, pipelined, efficient, extensible, stable, and easy-to-use.

Browse

Recent Submissions