• Progress in Medical Physics
• /
• v.30 no.2
• /
• pp.49-58
• /
• 2019

Purpose: Automated analytical systems have begun to emerge as a database system that enables the scanning of medical images to be performed on computers and the construction of big data. Deep-learning artificial intelligence (AI) architectures have been developed and applied to medical images, making high-precision diagnosis possible. Materials and Methods: For diagnosis, the medical images need to be labeled and standardized. After pre-processing the data and entering them into the deep-learning architecture, the final diagnosis results can be obtained quickly and accurately. To solve the problem of overfitting because of an insufficient amount of labeled data, data augmentation is performed through rotation, using left and right flips to artificially increase the amount of data. Because various deep-learning architectures have been developed and publicized over the past few years, the results of the diagnosis can be obtained by entering a medical image. Results: Classification and regression are performed by a supervised machine-learning method and clustering and generation are performed by an unsupervised machine-learning method. When the convolutional neural network (CNN) method is applied to the deep-learning layer, feature extraction can be used to classify diseases very efficiently and thus to diagnose various diseases. Conclusions: AI, using a deep-learning architecture, has expertise in medical image analysis of the nerves, retina, lungs, digital pathology, breast, heart, abdomen, and musculo-skeletal system.

• International Journal of Knowledge Content Development & Technology
• /
• v.7 no.4
• /
• pp.5-26
• /
• 2017

A cross-cultural comparison of social networking structure on McDonald's Facebook fan sites between Taiwan and the USA was conducted utilizing the individualism/collectivism dimension proposed by Hofstede. Four network indicators are used to describe the network structure of McDonald's Facebook fan sites: size, density, clique and centralization. Individuals who post on both Facebook sites for the year of 2012 were considered as network participants for the purpose of the study. Due to the huge amount of data, only one thread of postings was sampled from each month of the year of 2012. The final data consists of 1002 postings written by 896 individuals and 5962 postings written by 5532 individuals from Taiwan and the USA respectively. The results indicated that the USA McDonald's Facebook fan network has more fans, while Taiwan's McDonald's Facebook fan network is more densely connected. Cliques did form among the overall multiplex and within the individual uniplex networks in two countries, yet no significant differences were found between them. All the fan networks in both countries are relatively centralized, mostly on the site operators.

• International Journal of Computer Science & Network Security
• /
• v.22 no.12
• /
• pp.51-56
• /
• 2022

In space based constellation network, users are allowed to enter or leave the network arbitrarily. Hence, the number, identities and transmitted data of active users vary with time and have considerable impacts on the receiver's performance. The so-called problem of multiuser detection means identifying the identity of each active user and detecting the data transmitted by each active user. Traditional methods assume that the number of active users is equal to the maximum number of users that the network can hold. The model of traditional methods are simple and the performance are suboptimal. In this paper a Maximum A Posteriori Probability (MAP) based multiuser detection method is proposed. The proposed method models the activity state of users as Markov chain and transforms multiuser detection into searching optimal path in grid map with BCJR algorithm. Simulation results indicate that the proposed method obtains 2.6dB and 1dB E_b/N₀ gains respectively when activity detection error rate and symbol error rate reach 10^-3, comparing with reference methods.

• Proceedings of the KIEE Conference
• /
• 2006.04a
• /
• pp.273-275
• /
• 2006

In this paper, we introduce a new design of fuzzy neural networks using input-output data information of target system. The proposed fuzzy neural networks is constructed by input-output data information and used the center of data distance by HCM clustering to obtain the characteristics of data. A membership function is defined by HCM clustering and is applied input-output dat included each rule to conclusion polynomial functions. We use triangular membership functions and simplified fuzzy inference, linear fuzzy inference, and modified quadratic fuzzy inference in conclusion. In the networks learning, back propagation algorithm of network is used to update the parameters of the network. The proposed model is evaluated with benchmark data.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.61 no.2
• /
• pp.63-74
• /
• 2019

The objective of this study was to evaluate the influence of rainfall observation network on daily dam inflow using artificial neural networks(ANNs). Chungju Dam and Soyangriver Dam were selected for the study watershed. Rainfall and dam inflow data were collected as input data for construction of ANNs models. Five ANNs models, represented by Model 1 (In watershed, point rainfall), Model 2 (All in the Thiessen network, point rainfall), Model 3 (Out of watershed in the Thiessen network, point rainfall), Model 1-T (In watershed, area mean rainfall), Model 2-T (All in the Thiessen network, area mean rainfall), were adopted to evaluate the influence of rainfall observation network. As a result of the study, the models that used all station in the Thiessen network performed better than the models that used station only in the watershed or out of the watershed. The models that used point rainfall data performed better than the models that used area mean rainfall. Model 2 achieved the highest level of performance. The model performance for the ANNs model 2 in Chungju dam resulted in the $R^2$ value of 0.94, NSE of 0.94 $NSE_{ln}$ of 0.88 and PBIAS of -0.04 respectively. The model-2 predictions of Soyangriver Dam with the $R^2$ and NSE values greater than 0.94 were reasonably well agreed with the observations. The results of this study are expected to be used as a reference for rainfall data utilization in forecasting dam inflow using artificial neural networks.

• Annual Conference of KIPS
• /
• 2020.05a
• /
• pp.544-546
• /
• 2020

의료 데이터를 이용하여 인공지능 기계학습 연구를 수행할 때 자주 마주하는 문제는 데이터 불균형, 데이터 부족 등이며 특히 정제된 충분한 데이터를 구하기 힘들다는 것이 큰 문제이다. 본 연구에서는 이를 해결하기 위해 GAN(Generative Adversarial Network) 기반 고해상도 의료 영상을 생성하는 프레임워크를 개발하고자 한다. 각 해상도 마다 Scale 의 Gradient 를 동시에 학습하여 빠르게 고해상도 이미지를 생성해낼 수 있도록 했다. 고해상도 이미지를 생성하는 Neural Network 를 고안하였으며, PGGAN, Style-GAN 과의 성능 비교를 통해 제안된 모델이 양질의 고해상도 의료영상 이미지를 더 빠르게 생성할 수 있음을 확인하였다. 이를 통해 인공지능 기계학습 연구에 있어서 의료 영상의 데이터 부족, 데이터 불균형 문제를 해결할 수 있는 Data augmentation 이나, Anomaly detection 등의 연구에 적용할 수 있다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.3
• /
• pp.916-937
• /
• 2023

Most of the existing Distributed Denial-of-Service mitigation schemes in Software-Defined Networking are only implemented in the network domain managed by a single controller. In fact, the zombies for attackers to launch large-scale DDoS attacks are actually not in the same network domain. Therefore, abnormal traffic of DDoS attack will affect multiple paths and network domains. A single defense method is difficult to deal with large-scale DDoS attacks. The cooperative defense of multiple domains becomes an important means to effectively solve cross-domain DDoS attacks. We propose an efficient multi-domain DDoS cooperative defense mechanism by integrating blockchain and SDN architecture. It includes attack traceability, inter-domain information sharing and attack mitigation. In order to reduce the length of the marking path and shorten the traceability time, we propose an AS-level packet traceability method called ASPM. We propose an information sharing method across multiple domains based on blockchain and smart contract. It effectively solves the impact of DDoS illegal traffic on multiple domains. According to the traceability results, we designed a DDoS attack mitigation method by replacing the ACL list with the IP address black/gray list. The experimental results show that our ASPM traceability method requires less data packets, high traceability precision and low overhead. And blockchain-based inter-domain sharing scheme has low cost, high scalability and high security. Attack mitigation measures can prevent illegal data flow in a timely and efficient manner.

• Convergence Security Journal
• /
• v.12 no.1
• /
• pp.43-48
• /
• 2012

In this paper, the requirements for information security are presented in the ubiquitous ecological(u-Eco) city. The various definition of ubiquitous city is analyzed first, the concept of the u-Eco City, services and major projects are then presented. The framework of the integrated operating center for u-Eco city is proposed, the privacy, data security and network facility protection in the center are analyzed. Unlike to previously proposed security algorithms, the light-weight encoding algorithms(such as block/stream encoding, pseudo-random generator, hash function, and public key encoding) in the u-Eco city center are required to communicate the information in the ubiquitous sensor network. Furthermore, the principal policies guaranteeing the secrecy and authentication for the private information are also presented.

• Proceedings of the IEEK Conference
• /
• 2002.07a
• /
• pp.30-33
• /
• 2002

To obtain good performance of radial basis function (RBF) neural networks, it needs very careful consideration in design. The selection of several parameters such as the number of centers and widths of the radial basis functions must be considered carefully since they critically affect the network's performance. We propose a learning algorithm for growing of complexity of RBF neural networks which is adapted automatically according to the complexity of tasks. The algorithm generates a new basis function based on the errors of network, the percentage of decreasing rate of errors and the nearest distance from input data to the center of hidden unit. The RBF's center is located at the point where the maximum of absolute interference error occurs in the input space. The width is calculated based on the standard deviation of distance between the center and inputs data. The steepest descent method is also applied for adjusting the weights, centers, and widths. To demonstrate the performance of the proposed algorithm, general problem of function estimation is evaluated. The results obtained from the simulation show that the proposed algorithm for RBF neural networks yields good performance in terms of convergence and accuracy compared with those obtained by conventional multilayer feedforward networks.

• Journal of Korea Water Resources Association
• /
• v.41 no.10
• /
• pp.1059-1065
• /
• 2008

Rainfall data is the most basic input data to analyze the hydrological phenomena and can be missing due to various reasons. In this research, a neural network based model to estimate missing rainfall data as approximate values was developed for 12 rainfall stations in the Soyang river basin to improve existing methods. This approach using neural network has shown to be useful in many applications to deal with complicated natural phenomena and displayed better results compared to the popular offline estimating methods, such as RDS(Reciprocal Distance Squared) method and AMM(Arithmetic Mean Method). Additionally, we proposed automated data reconciliation systems composed of a neural network learning processer to be capable of real-time reconciliation to transmit reliable hydrological data online.