Article
Computer Science, Information Systems
Gihan J. Mendis, Yifu Wu, Jin Wei, Moein Sabounchi, Rigoberto Roche
Summary: The proposed decentralized, secure, and privacy-preserving computing paradigm enables an asynchronized cooperative computing process amongst distributed and untrustworthy computing nodes. This paradigm is designed by exploring blockchain, decentralized learning, and homomorphic encryption techniques.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTING
(2021)
Article
Computer Science, Information Systems
Jianling Gao, Nan Zhao, Ning Wang, Shuang Hao, Haoyan Wu
Summary: Choosing useful indexes for a relational database is crucial for query optimization. However, existing methods suffer from inaccurate cost estimates and disregard the relationships among indexes, leading to suboptimal solutions and unnecessary training costs. To address these issues, we propose DeepIndex, an automatic index selector with a learning-based cost estimator that accurately predicts the benefits of indexes and considers their relationships. Experimental results demonstrate that our model outperforms state-of-the-art approaches in terms of storage costs and relative execution costs.
INFORMATION SCIENCES
(2022)
Article
Chemistry, Analytical
Seemab Khan, Muhammad Attique Khan, Majed Alhaisoni, Usman Tariq, Hwan-Seung Yong, Ammar Armghan, Fayadh Alenezi
Summary: Human action recognition (HAR) is crucial for smart surveillance systems but poses challenges due to the variety of actions and large video sequences. Deep learning (DL) systems have shown significant success in HAR, achieving high accuracies on multiple datasets. The proposed DL-based design includes feature mapping, fusion, selection steps, and outperforms state-of-the-art methods in terms of computational time.
Article
Computer Science, Information Systems
Ti Ti Nguyen, Kim-Khoa Nguyen
Summary: This paper proposes a novel learning framework using beam-steering technique to estimate signal strength from the base station to the user in order to determine the suitable beam for a specific user and minimize the transmit power cost. The missing data problem is addressed and the long-short term memory (LSTM) is employed to select the suitable beam. The proposed learning framework outperforms state-of-the-art prediction strategies and approximates the best performance.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2023)
Article
Engineering, Multidisciplinary
Yuequan Bao, Hui Li
Summary: The conventional vibration-based methods face challenges in accurately detecting structural damages, thus necessitating the development of novel diagnosis and prognosis methods based on various monitoring data.
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2021)
Article
Computer Science, Information Systems
Meng Shen, Ke Ye, Xingtong Liu, Liehuang Zhu, Jiawen Kang, Shui Yu, Qi Li, Ke Xu
Summary: Traffic analysis is the process of monitoring network activities and extracting valuable information. With the encryption of network traffic, traditional analysis methods become less effective, and machine learning has emerged as a powerful tool for encrypted traffic analysis. This paper presents a comprehensive survey on recent achievements in machine learning-powered encrypted traffic analysis, reviewing the literature, abstracting the workflow, and discussing future directions.
IEEE COMMUNICATIONS SURVEYS AND TUTORIALS
(2023)
Article
Computer Science, Information Systems
Yoones Rezaei, Talha Khan, Stephen Lee, Daniel Mosse
Summary: Advances in deep vision techniques and smart cameras will drive the next generation of video analytics. However, video analytics applications consume large amounts of energy as deep learning techniques are power-hungry. This article proposes RL-CamSleep, a deep reinforcement learning-based technique, to reduce energy consumption by activating cameras only when necessary.
ACM TRANSACTIONS ON SENSOR NETWORKS
(2023)
Article
Multidisciplinary Sciences
Siti Nurmaini, Annisa Darmawahyuni, Muhammad Naufal Rachmatullah, Firdaus Firdaus, Ade Iriani Sapitri, Bambang Tutuko, Alexander Edo Tondas, Muhammad Hafizh Permana Putra, Anggun Islami
Summary: This paper proposes a robust delineation model based on a convolutional recurrent network with grid search optimization for the accurate classification and interpretation of ECG waveform. By generating and evaluating 36 models, high accuracy and performance are achieved on multiple datasets. Furthermore, the method is capable of handling noise interference, baseline drift, and abnormal morphology, demonstrating good robustness.
SCIENTIFIC REPORTS
(2023)
Article
Engineering, Biomedical
Zhen Wang, Jiajia Wang, Xiang Shi, Zhengfeng Zhu, Peining Chen, Huisheng Peng
Summary: Learning is essential for human growth, but traditional methods are insufficient for the abundant knowledge explosion. To improve efficiency, electronic neurons (E-neurons) inspired by cloud storage are proposed to transfer knowledge to humans without training. Feasibility of this concept is demonstrated through experiments.
ADVANCED HEALTHCARE MATERIALS
(2023)
Article
Microbiology
Xiaodi Yang, Shiping Yang, Panyu Ren, Stefan Wuchty, Ziding Zhang
Summary: Identifying human-virus protein-protein interactions (PPIs) is crucial for understanding viral infection mechanisms and antiviral response. Deep learning, as a promising method, has shown great potential in predicting human-virus PPIs. This review focuses on recent advances in deep learning-powered human-virus PPI predictions, discussing technical details, challenges, and future perspectives.
FRONTIERS IN MICROBIOLOGY
(2022)
Article
Computer Science, Theory & Methods
Muhammad Attique Khan, Asif Mehmood, Seifedine Kadry, Nouf Abdullah Almujally, Majed Alhaisoni, Jamel Balili, Abdullah Al Hejaili, Abed Alanazi, Shtwai Alsubai, Abdullah Alqatani
Summary: Industrial advancements and financial stakes drive the aims of smart cities, which focus on increasing efficiency and citizens' quality of life. Human Gait Recognition (HGR) is an important application that utilizes walking styles for individual recognition. This paper proposes a deep learning method for HGR using a large gait database and transfer learning, achieving high accuracy by addressing constraints such as poor lighting and varying angles. The proposed technique combines improved BAT algorithm, entropy selection, and canonical correlation analysis to extract and fuse features for final gait recognition using a softmax classifier.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Cell Biology
Edo Cohen-Karlik, Zamzam Awida, Ayelet Bergman, Shahar Eshed, Omer Nestor, Michelle Kadashev, Sapir Ben Yosef, Hussam Saed, Yishay Mansour, Amir Globerson, Drorit Neumann, Yankel Gabet
Summary: The current manual methods used to study osteoclast differentiation are cumbersome, time-consuming, and subject to operator bias. Therefore, replacing this with a completely automated process using computer vision algorithms can improve efficiency and reduce human bias.
FRONTIERS IN CELL AND DEVELOPMENTAL BIOLOGY
(2021)
Article
Computer Science, Information Systems
Iram Bibi, Adnan Akhunzada, Neeraj Kumar
Summary: Distributed Industrial Internet of Things (IIoT) has revolutionized the industrial sector, but threat hunting and intelligence in distributed IIoT is complex due to lack of standard architectures. The authors propose a self-learning multivector threat intelligence and detection mechanism to defend IIoT systems. They introduce a novel ConvLSTM2D mechanism that can efficiently tackle dynamic variants of emerging IIoT threats. The proposed mechanism outperforms benchmark algorithms in detection accuracy with minimal tradeoff in speed efficiency.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Review
Computer Science, Information Systems
Asma Alotaibi, Ahmed Barnawi
Summary: The Internet of things (IoT) is a rapidly developing technology that enables smart services in various domains. Securing 6G/massive IoT networks against threats, especially novel attacks, is a major challenge. Innovative architectures and paradigms empowered by artificial intelligence and key networking enablers are urgently needed. Researchers are using machine learning and deep learning techniques to improve cyber threat detection. Designing an intrusion detection system (IDS) for massive IoT applications is a challenge that requires consideration of multiple factors. This survey provides a comprehensive study on massive IoT security aspects, particularly IDS systems, in the context of 6G networks.
INTERNET OF THINGS
(2023)
Article
Engineering, Multidisciplinary
Yuchen Mu, Navneet Garg, Tharmalingam Ratnarajah
Summary: This paper investigates the impact of imperfections in both uplink and downlink links due to the limited reliability of wireless channels in federated learning. It is found that the reliability of downlink links is more critical than that of uplink links, and varying the transmit power during training can reduce energy consumption.
IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING
(2022)
Article
Nanoscience & Nanotechnology
Shunrui Luo, Kostiantyn Turcheniuk, Lihua Chen, Ah-Young Song, Wenqiang Hu, Xiaolei Ren, Zifei Sun, Rampi Ramprasad, Gleb Yushin
Summary: We report a new synthesis pathway for Mg n-propoxide nanowires from Mg ethoxide nanoparticles. The morphology transformation from nanoparticles to nanowires was studied using characterization techniques such as SEM, FTIR, and NMR spectroscopy. The ligand exchange and increased fraction of OH groups greatly enhanced Mg alkoxide bonding and facilitated the formation and growth of the Mg n-propoxide nanowires.
ACS APPLIED MATERIALS & INTERFACES
(2022)
Article
Chemistry, Physical
R. Datta, R. Ramprasad, S. Venkatram
Summary: This study utilizes a deep neural network to rapidly and accurately predict the conductivity of ionic liquids (ILs) and identifies key chemical structural characteristics that correlate with the ionic conductivity. The findings provide guidance for the design and synthesis of new highly conductive ILs.
JOURNAL OF CHEMICAL PHYSICS
(2022)
Article
Chemistry, Physical
Harikrishna Sahu, Kuan-Hsuan Shen, Joseph H. Montoya, Huan Tran, Rampi Ramprasad
Summary: Researchers have developed a Python toolkit called PSP for generating polymer models based on SMILES strings. Users can adjust parameters to manage the quality and scale of models, with output structures and forcefield parameter files available for downstream simulations. The PSP package also includes a Colab notebook for user interaction and learning.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2022)
Article
Chemistry, Physical
Abdullah Alamri, Chao Wu, Ankit Mishra, Lihua Chen, Zongze Li, Ajinkya Deshmukh, Jierui Zhou, Omer Yassin, Rampi Ramprasad, Priya Vashishta, Yang Cao, Gregory Sotzing
Summary: Traditionally, polymers with good thermal stability have low charge-discharge efficiency under high electric field and elevated temperature. By modifying the molecular structure, we have successfully optimized the dielectric properties of polyetherimide and improved its charge-discharge efficiency, making it a potential candidate for high-temperature energy storage applications.
CHEMISTRY OF MATERIALS
(2022)
Article
Chemistry, Physical
Yifan Liu, Emily K. McGuinness, Benjamin C. Jean, Yi Li, Yi Ren, Beatriz G. del Rio, Ryan P. Lively, Mark D. Losego, Rampi Ramprasad
Summary: This paper investigates the chemical reaction pathways and the chemical structure of hybrid membranes during the vapor-phase infiltration process. The results show that a stable coordination is formed between the metal-organic precursor and PIM-1 during the precursor exposure step, and subsequent water vapor exposure leads to the formation of the final hybrid membrane through a series of exothermic reactions.
JOURNAL OF PHYSICAL CHEMISTRY B
(2022)
Article
Polymer Science
Kellie A. Stellmach, McKinley K. Paul, Mizhi Xu, Yong-Liang Su, Liangbing Fu, Aubrey R. Toland, Huan Tran, Lihua Chen, Rampi Ramprasad, Will R. Gutekunst
Summary: This report investigates the polymerization thermodynamics of thiolactone monomers and explores the effects of substitution patterns and sulfur heteroatom incorporation. Computational studies reveal the significance of conformation in modulating the enthalpy of polymerization, enabling high conversion rates at near-ambient temperatures.
Article
Engineering, Chemical
Shubham Jamdade, Rishi Gurnani, Hanjun Fang, Salah Eddine Boulfelfel, Rampi Ramprasad, David S. Sholl
Summary: We developed a computational approach to screen MOFs for oxygen-helium separation at low temperatures. Through molecular simulations and stability evaluation, we identified high-performance materials for this separation process. This method can also be applied to selecting adsorbents for other gas separations.
INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH
(2023)
Article
Chemistry, Physical
Huan Tran, Kuan-Hsuan Shen, Shivank Shukla, Ha-Kyung Kwon, Rampi Ramprasad
Summary: Modern fuel cell technologies use Nafion for proton-exchange membrane and as the binding material for the catalyst layers. This study proposes an informatics-based scheme to search large polymer chemical spaces and identifies 60 new polymer candidates for various applications in fuel cells.
JOURNAL OF PHYSICAL CHEMISTRY C
(2023)
Article
Chemistry, Physical
Rishi Gurnani, Christopher Kuenneth, Aubrey Toland, Rampi Ramprasad
Summary: Artificial intelligence-based methods are effective in screening polymer libraries for experimental inquiry. Our approach uses machine learning to extract important features from polymer repeat units, speeding up feature extraction by 1-2 orders of magnitude without compromising model accuracy. This approach will enable more sophisticated and large-scale screening technologies in polymer informatics.
CHEMISTRY OF MATERIALS
(2023)
Article
Chemistry, Multidisciplinary
Yong-Liang Su, Liang Yue, Huan Tran, Mizhi Xu, Anthony Engler, Rampi Ramprasad, H. Jerry Qi, Will R. R. Gutekunst
Summary: In this study, a chemically recyclable polythioether system based on nucleophilic aromatic substitution (SNAr) was developed. The system demonstrated chain-growth ring-opening polymerization through SNAr reactions, with fast reaction rates and efficient polymerization and depolymerization cycles. The resulting polythioether materials showed comparable performance to commercial thermoplastics, and could be depolymerized to the original monomers in high yields.
JOURNAL OF THE AMERICAN CHEMICAL SOCIETY
(2023)
Article
Polymer Science
Shruti Venkatram, Jena McCollum, Natalie Stingelin, Blair Brettmann
Summary: In the polymer field, it is crucial to differentiate between the degree of crystallinity and the crystalline quality. These structural features have a significant impact on the properties of plastic materials and determine various processes in semiconducting polymers. Therefore, it is important to establish clear correlations between structure, processing, and properties, attributing specific functions to the degree of crystallinity and/or crystalline quality. This article discusses the challenges of identifying and distinguishing these structural characteristics using commonly applied measuring techniques and theoretical approaches.
POLYMER INTERNATIONAL
(2023)
Article
Chemistry, Physical
Pranav Shetty, Arunkumar Chitteth Rajan, Chris Kuenneth, Sonakshi Gupta, Lakshmi Prerana Panchumarti, Lauren Holm, Chao Zhang, Rampi Ramprasad
Summary: This study used natural language processing methods to extract material property data from polymer literature abstracts. By training the MaterialsBERT language model, we obtained around 300,000 material property records for analysis in various applications such as fuel cells, supercapacitors, and polymer solar cells.
NPJ COMPUTATIONAL MATERIALS
(2023)
Review
Materials Science, Multidisciplinary
Janhavi Nistane, Lihua Chen, Youngjoo Lee, Ryan Lively, Rampi Ramprasad
Summary: This study presents a machine learning model that can instantly predict the temperature-dependent Flory-Huggins interaction parameter for polymer-solvent mixtures. The model has been trained using a large dataset of experimental data and demonstrates high accuracy and generality.
MRS COMMUNICATIONS
(2022)
Article
Chemistry, Physical
Huan Tran, Aubrey Toland, Kellie Stellmach, McKinley K. Paul, Will Gutekunst, Rampi Ramprasad
Summary: Researchers have developed a first-principles computational scheme to calculate Delta H-ROP for polymer systems, achieving a root-mean-square error of 7 kJ/mol on a benchmark set of 42 ROP polymers. This development paves the way for building a high-quality database of Delta H-ROP for predictive machine-learning models and accelerating the design of depolymerizable polymers.
JOURNAL OF PHYSICAL CHEMISTRY LETTERS
(2022)