http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Required Base Station Density for Communication-Efficient Federated Learning in Small-Cell Networks
Khoa Anh Nguyen,Quan Anh Nguyen,Jun-Pyo Hong 한국통신학회 2021 한국통신학회 학술대회논문집 Vol.2021 No.6
Federated Learning (FL) generates massive benefits of shared machine learning models without violating security and privacy requirements, making this setting relevant for many wireless applications. In this paper, we comprehensively investigate the effects of geographic node deployment on the model aggregation in federated learning on the basis of stochastic geometry-based analysis. Based on the coverage probability expression derived with stochastic geometry framework, we derive and discuss the minimum required base station density for achieving a target model aggregation rate in small-cell networks with algorithms for optimizing the target transmission rate and the base station density.
Khoa, H.X.,Tuan, N.Q.,Lee, Y.H.,Lee, B.H.,Viet, N.H.,Kim, J.S. The Korean Powder Metallurgy Institute 2013 한국분말재료학회지 (KPMI) Vol.20 No.3
$TiB_2$-reinforced iron matrix composite (Fe-$TiB_2$) powder was in-situ fabricated from titanium hydride ($TiH_2$) and iron boride (FeB) powders by the mechanical activation and a subsequent reaction. Phase formation of the composite powder was identified by X-ray diffraction (XRD). The morphology and phase composition were observed and measured by field emission-scanning electron microscopy (FE-SEM) and energy-dispersive X-ray spectroscopy (EDS), respectively. The results showed that $TiB_2$ particles formed in nanoscale were uniformly distributed in Fe matrix. $Fe_2B$ phase existed due to an incomplete reaction of Ti and FeB. Effect of milling process and synthesis temperature on the formation of composite were discussed.
Memristor Binarized Neural Networks
Khoa Van Pham,Tien Van Nguyen,Son Bao Tran,HyunKyung Nam,Mi Jung Lee,Byung Joon Choi,Son Ngoc Truong,Kyeong-Sik Min 대한전자공학회 2018 Journal of semiconductor technology and science Vol.18 No.5
Binarized Neural Networks (BNNs) use only binary synapses of +1 and -1, not allowing any intermediate weights between -1 and +1. Though the recognition rate of BNNs is lower than the conventional Deep Neural Networks (DNNs), BNNs have attracted many interests nowadays, because BNNs do not need the complicated multiplication such as DNNs. Binary memristor crossbars can be very suitable to realize BNN hardware. This is because, in memristor BNNs, simple binary operation can be performed in bitwise manner for all the columns in memristor crossbars, simultaneously. In this paper, single-column and double-column memristor BNNs are presented, respectively. In addition, ReLU and sigmoid activation function circuits are also proposed with CMOS circuits. The designed Memristor-CMOS hybrid circuits of BNNs have been tested for MNIST vectors. The memristor BNNs could recognize almost 90% MNIST digits when the memristance variation is as large as 25%. For variation tolerance, the memristor BNNs are compared with the multi-valued memristor neural networks such as 4-bit, 6-bit, etc, in this paper. As a result, it has been confirmed the memristor BNNs become more variation-tolerant than the multi-valued memristor NNs when the variation becomes larger than 22%. Comparing the single-column and doublecolumn BNNs in this paper indicates that the singlecolumn BNN can save power consumption and array area almost by half than the double-column. This is because the single-column has just half memristors than the double-column. And, we measured the single-column and double-column BNNs using the fabricated memristor array. In this measurement, both the double-column and single-column BNNs were observed to work well.
Near-Optimal Deviation-Proof Medium Access Control Designs in Wireless Networks
Khoa Tran Phan,Jaeok Park,van der Schaar, M. IEEE 2012 IEEE/ACM transactions on networking Vol.20 No.5
<P>Distributed medium access control (MAC) protocols are essential for the proliferation of low-cost, decentralized wireless local area networks (WLANs). Most MAC protocols are designed with the presumption that nodes comply with prescribed rules. However, selfish nodes have natural motives to manipulate protocols in order to improve their own performance. This often degrades the performance of other nodes as well as that of the overall system. In this paper, we propose a class of protocols that limit the performance gain from selfish manipulation while incurring only a small efficiency loss. The proposed protocols are based on the idea of a review strategy, with which nodes collect signals about the actions of other nodes over a period of time, use a statistical test to infer whether or not other nodes are following the prescribed behavior, and trigger a punishment if a deviation is inferred. We consider the cases of private and public signals and provide analytical and numerical results to demonstrate the properties of the proposed protocols.</P>
Khoa, Nguyen Tri,Kim, Soon Wook,Yoo, Dae-Hwang,Cho, Shinuk,Kim, Eui Jung,Hahn, Sung Hong American Chemical Society 2015 ACS APPLIED MATERIALS & INTERFACES Vol.7 No.6
<P>Heterostructures of gold-nanoparticle-decorated reduced-graphene-oxide (rGO)-wrapped ZnO hollow spheres (Au/rGO/ZnO) are synthesized using tetra-<I>n</I>-butylammonium bromide as a mediating agent. The structure of amorphous ZnO hollow spheres is found to be transformed from nanosheet- to nanoparticle-assembled hollow spheres (nPAHS) upon annealing at 500 °C. The ZnO nPAHS hybrids with Au/rGO are characterized using various techniques, including photoluminescence, steady-state absorbance, time-resolved photoluminescence, and photocatalysis. The charge-transfer time of ZnO nPAHS is found to be 87 ps, which is much shorter than that of a nanorod (128 ps), nanoparticle (150 ps), and nanowall (990 ps) due to its unique structure. The Au/rGO/ZnO hybrid shows a higher charge-transfer efficiency of 68.0% in comparison with rGO/ZnO (40.3%) and previously reported ZnO hybrids. The photocatalytic activities of the samples are evaluated by photodegrading methylene blue under black-light irradiation. The Au/rGO/ZnO exhibits excellent photocatalytic efficiency due to reduced electron–hole recombination, fast electron-transfer rate, and high charge-transfer efficiency.</P><P><B>Graphic Abstract</B> <IMG SRC='http://pubs.acs.org/appl/literatum/publisher/achs/journals/content/aamick/2015/aamick.2015.7.issue-6/acsami.5b00152/production/images/medium/am-2015-00152a_0004.gif'></P><P><A href='http://pubs.acs.org/doi/suppl/10.1021/am5b00152'>ACS Electronic Supporting Info</A></P>
KHOA, Bui Thanh Korea Distribution Science Association 2020 The Journal of Asian Finance, Economics and Busine Vol.7 No.2
Customer loyalty is a desired and essential result for companies in a business environment with a lot of competition, of which the fashion industry is no exception. For the fashion industry, the competition between businesses is increasingly fierce with the appearance of tailored fashion brands for each customer, also known as a designed fashion brand. The objective of this study is to examine the relationship between the antecedents of the relationship (reputation of designer, social media marketing) and customer relationship (trust and intimacy), and loyalty of customer for a designed fashion brand. The mixed research method is used to achieve research objectives with qualitative and quantitative research method. Qualitative data was collected through in-depth interview with 12 participants in Ho Chi Minh City, while quantitative data was collected by a survey of 799 respondents in Ho Chi Minh City and Hanoi City. The results show that the antecedents of relationship marketing positively affect on customer relationship and customer loyalty. Moreover, customer relationship has a positive impact on the customer loyalty when shopping for fashion products. From the research results, some managerial implications are also proposed for the designed fashion brands to improve the relationship marketing performance.