Federated learning is a type of distributed machine learning in which models learn by using large-scale decentralized data between servers and devices. In a short-range wireless communication ...environment, it can be difficult to apply federated learning because the number of devices in one access point (AP) is small, which can be small enough to perform federated learning. Therefore, it means that the minimum number of devices required to perform federated learning cannot be matched by the devices included in one AP environment. To do this, we propose to obtain a uniform global model regardless of data distribution by considering the multi-AP coordination characteristics of IEEE 802.11be in a decentralized federated learning environment. The proposed method can solve the imbalance in data transmission due to the non-independent and identically distributed (non-IID) environment in a decentralized federated learning environment. In addition, we can also ensure the fairness of multi-APs and determine the update criteria for newly elected primary-APs by considering the learning training time of multi-APs and energy consumption of grouped devices performing federated learning. Thus, our proposed method can determine the primary-AP according to the number of devices participating in the federated learning in each AP during the initial federated learning to consider the communication efficiency. After the initial federated learning, fairness can be guaranteed by determining the primary-AP through the training time of each AP. As a result of performing decentralized federated learning using the MNIST and FMNIST dataset, the proposed method showed up to a 97.6% prediction accuracy. In other words, it can be seen that, even in a non-IID multi-AP environment, the update of the global model for federated learning is performed fairly.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
Federated learning is a distributed machine learning framework that enables a large number of devices to cooperatively train a model without data sharing. However, because federated learning trains a ...model using non-independent and identically distributed (non-IID) data stored at local devices, the weight divergence causes a performance loss. This paper focuses on solving the non-IID problems and proposes Kalman filter-based clustering federated learning method called K-FL to get performance gain by providing a specific model with low variance to the device. To the best of our knowledge, it is the first clustering federated learning method that can train a model requiring fewer communication rounds under the premise that non-IID environment without any prior knowledge and an initial value set by the user. From simulations, we demonstrate that the proposed K-FL can train a model much faster, requiring fewer communication rounds than FedAvg and LG-FedAvg when testing neural networks using the MNIST, FMNIST, and CIFAR-10 datasets. As a numerical result, it is shown that the accuracy is improved in all datasets while the computational time cost is reduced by <inline-formula> <tex-math notation="LaTeX">1.43\times </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">1.67\times </tex-math></inline-formula>, and <inline-formula> <tex-math notation="LaTeX">1.63\times </tex-math></inline-formula> compared to FedAvg, respectively.
Abstract
Mussel periostracum, a nonliving multifunctional gel that covers the rigid inorganic shells of mussels, provides protection against mechanical impacts, biofouling, and corrosion in harsh ...ocean environments. The inner part of the periostracum, which emerges from biological tissues, functions as a natural interface between tissue and inorganic materials. The periostracum shows significant potential for application in implantable devices that provide interfaces; however, this system remains unexplored. In this study, we revealed that the inner periostracum performs graded mechanical functions and efficiently dissipates energy to accommodate differences in stiffness and stress types on both sides. On the tissue end, the lightly pigmented periostracum exhibits extensibility and energy dissipation under repetitive tension. This process was facilitated by the slipping and reassembly of β-strands in the discovered major proteins, which we named periostracin proteins. On the shell end, the highly pigmented, mineralized, and porous segment of the periostracum provided stiffness and cushioned against compressive stresses exerted by the shell valves during closure. These findings offer a novel possibilities for the design of interfaces that bridge human tissue and devices.
Complex coacervation is an emerging liquid/liquid phase separation (LLPS) phenomenon that behaves as a membrane-less organelle in living cells. Yet while one of the critical factors for complex ...coacervation is temperature, little analysis and research has been devoted to the temperature effect on complex coacervation. Here, we performed a complex coacervation of cationic protamine and multivalent anions (citrate and tripolyphosphate (TPP)). Both mixtures (i.e., protamine/citrate and protamine/TPP) underwent coacervation in an aqueous solution, while a mixture of protamine and sodium chloride did not. Interestingly, the complex coacervation of protamine and multivalent anions showed upper critical solution temperature (UCST) behavior, and the coacervation of protamine and multivalent anions was reversible with solution temperature changes. The large asymmetry in molecular weight between positively charged protamine (~4 kDa) and the multivalent anions (<0.4 kDa) and strong electrostatic interactions between positively charged guanidine residues in protamine and multivalent anions were likely to contribute to UCST behavior in this coacervation system.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
The biofouling of marine organisms on a surface induces serious economic damage. One of the conventional anti-biofouling strategies is the use of toxic chemicals. In this study, a new eco-friendly ...oleamide–PDMS copolymer (OPC) is proposed for sustainable anti-biofouling and effective drag reduction. The anti-biofouling characteristics of the OPC are investigated using algal spores and mussels. The proposed OPC is found to inhibit the adhesion of algal spores and mussels. The slippery features of the fabricated OPC surfaces are examined by direct measurement of pressure drops in channel flows. The proposed OPC surface would be utilized in various industrial applications including marine vehicles and biomedical devices.
Full text
Available for:
IJS, KILJ, NUK, PNG, UL, UM, UPUK
표준화된 곰피추출물의 항산화 활성 및 콜레스테롤 개선 효과 한웅호; 김우혁; 최선일 ...
Han'gug sigpum wi'saeng anjeonseong haghoeji,
08/2021, Volume:
36, Issue:
4
Journal Article
Peer reviewed
Open access
Ecklonia stolonifera, which belongs to the family Laminariaceae, is an edible perennial brown marine alga that is widely distributed, and is rich in polyphenols, including dieckol. Here, we ...investigated the radical scavenging activities of E. stolonifera extract (ESE) using various in vitro models. We further evaluated the effect of ESE on the cholesterol secretion inhibition activity in HepG2 cells, as well as the hydroxymethylglutaryl-coenzyme A (HMG-CoA) reductase activity. Our results showed that the total phenol, total flavonoid, and dieckol contents of ESE were 9.64±0.04 mg GAE/g, 2.72±0.08 mg RE/g and 27.42±0.66 mg/g, respectively. The antioxidant activity of ESE increased in a dose-dependent manner. In addition, the ESE inhibited cholesterol secretion from HepG2 cells with anti-HMG-CoA reductase activity. These results suggested that ESE possesses antioxidant and anti-cholesterol activities, and can therefore be used as a preclinical bioresource for development of health functional foods. 본 연구에서는 표준화된 곰피추출물의 항산화 및 콜레스테롤 개선에 대한 효능평가를 통해 건강기능식품 소재로서의 가치를 검토하기 위해 총 폴리페놀, 총 플라보노이드 및 dieckol 함량을 측정하였으며 DPPH, ABTS radical 소거능, reducing power 및 FRAP 활성을 통하여 곰피추출물의 in vitro 항산화 활성을 조사하였고 표준화된 곰피추출물의 HMG-CoA reductase 저해 활성 및 세포 내 콜레스테롤 생성 억제 효능을 평가하였다. 표준화된 곰피추출물의 총 폴리페놀, 총 플라보노이드 및 dieckol 함량은 각각 9.64±0.04 mg GAE/g, 2.72±0.08 mg RE/g, 27.42±0.66 mg/g으로 나타났다. 표준화된 곰피추출물의 in vitro 항산화활성, HMG-CoA reductase 저해활성 및 세포 내 콜레스테롤 생성 억제 효능은 농도의존적으로 증가하는 경향을 보였으며 이는 표준화된 곰피추출물에 함유되어 있는 페놀성 화합물에 기인된 효능으로 사료되며 항산화성분, 항산화 효과, 콜레스테를 개선 효능간의 상관관계가 있음을 확인하였다. 향후, 표준화된 곰피추출물에 대한 in vivo 모델에서의 전임상 연구 및 작용기전 입증되면 인체적용시험을 통해 이중기능성을 갖는 건강기능식품의 개발이 가능할 것으로 사료된다.
The load-bearing proteins in mussel holdfasts rely on condensed tris-catecholato-Fe3+ coordination complexes for their toughness and shock-absorbing properties, and this feature has been successfully ...translated into synthetic materials with short-term high-performance properties. However, oxidation of catecholic DOPA (3,4-dihydroxyphenylalanine) remains a critical impediment to achieving materials with longer-lasting performance. Here, following the natural mussel pathway for protein processing, we explore how DOPA oxidation impacts coacervation of mussel foot protein-1 (mfp-1) and its capacity for phase-specific metal uptake in vitro. Without metal, DOPA oxidation changed the rheological properties (i.e., viscosity, loss, and storage moduli) of mfp-1 coacervate droplets. However, oxidation-dependent changes were recovered with dithiothreitol (DTT), completely restoring the behavior of mfp-1 coacervates prior to oxidation. With metal, mfp-1 coacervates exhibited gel-like behavior with high viscosity and cohesive forces by forming recognizable bis- and tris-catecholato-Fe complexes, linked to increased energy dissipation and toughness of byssus. These results indicate that Fe3+-mediated conversion of liquid–liquid phase-separated polymers into metal-coordinated networks is thorough and rapid, and DTT effectively maintains redox integrity. Our study provides much-needed improvements for processing catechol-functionalized polymers into high-performance materials.
Full text
Available for:
IJS, KILJ, NUK, PNG, UL, UM
Biological organisms produce high-performance composite materials, such as bone, wood and insect cuticle, which provide inspiration for the design of novel materials. Ascidians (sea squirts) produce ...an organic exoskeleton, known as a tunic, which has been studied quite extensively in several species. However, currently, there are still gaps in our knowledge about the detailed structure and composition of this cellulosic biocomposite. Here, we investigate the composition and hierarchical structure of the tough tunic from the species Halocynthia roretzi, through a cross-disciplinary approach combining traditional histology, immunohistochemistry, vibrational spectroscopy, X-ray diffraction, and atomic force and electron microscopies. The picture emerging is that the tunic of H. roretzi is a hierarchically-structured composite of cellulose and proteins with several compositionally and structurally distinct zones. At the surface is a thin sclerotized cuticular layer with elevated composition of protein containing halogenated amino acids and cross-linked via dityrosine linkages. The fibrous layer makes up the bulk of the tunic and is comprised primarily of helicoidally-ordered crystalline cellulose fibres with a lower protein content. The subcuticular zone directly beneath the surface contains much less organized cellulose fibres. Given current efforts to utilize biorenewable cellulose sources for the sustainable production of bio-inspired composites, these insights establish the tunic of H. roretzi as an exciting new archetype for extracting relevant design principles.
Tunicates are the only animals able to produce cellulose. They use this structural polysaccharide to build an exoskeleton called a tunic. Here, we investigate the composition and hierarchical structure of the tough tunic from the sea pineapple Halocynthia roretzi through a multiscale cross-disciplinary approach. The tunic of this species is a composite of cellulose and proteins with two distinct layers. At the surface is a thin sclerotized cuticular layer with a higher protein content containing halogenated amino acids and cross-linked via dityrosine linkages. The fibrous layer makes up the bulk of the tunic and is comprised of well-ordered cellulose fibres with a lower protein content. Given current efforts to utilize cellulose to produce advanced materials, the tunic of the sea pineapple provides a striking model for the design of bio-inspired cellulosic composites.
Display omitted
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
Federated Learning (FL) has a different learning framework from existing machine learning, which had to centralize training data. Federated learning has the advantage of protecting privacy because ...learning is performed on each client device rather than the central server, and only the weight parameter values, which are the learning results, are sent to the central server. However, the performance of federated learning shows relatively low performance compared to cloud computing, and in reality, it is difficult to build a federated learning environment due to the high communication cost between the server and multiple clients. In this paper, we propose Federated Learning with Clustering algorithms (FLC). The proposed FLC is a method of clustering clients with similar characteristics by analyzing the weights of each layer of a machine learning model, and performing federated learning among the clustered clients. The proposed FLC can reduce the communication cost for each model by reducing the number of clients corresponding to each model. As a result of extensive simulation, it is confirmed that the accuracy is improved by 2.4% and the loss by 47% through the proposed FLC compared to the standard federated learning.
Vehicle automation is expected to reduce the cost of driving and to solve social problems, such as traffic accidents. However, despite the expected benefits of autonomous vehicles, some consumers ...hesitate to adopt autonomous vehicles due to fears of negative effects from vehicle automation. This study analyzes consumers' heterogeneous preferences for autonomous vehicles using a discrete choice experiment and a latent class model, which enables the clustering of consumer preferences. As a result, consumers' preferences for autonomous vehicles are largely divided into two groups: one group is technology-friendly users (78.0%), and the other group is change avoiders (22.0%). Technology-friendly users tend to prefer advanced vehicle automation and intend to offer their personal information for advanced functions if the information is highly secured and protected. However, change avoiders express tendencies to avoid highly active vehicle automation and are reluctant to use their personal information for operating automation systems. Change avoiders think that laws or regulations specialised for vehicle automation are redundant. In addition, technology-friendly users and change avoiders prefer conditional automation and driver assistance, respectively.
Full text
Available for:
BFBNIB, NUK, PILJ, SAZU, UL, UM, UPUK
PDF
