• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2020년도 학술발표회
• /
• pp.170-170
• /
• 2020

Due to the development of technologies, complex computation of huge data set is possible with a prevalent personal computer. Therefore, machine learning methods have been widely applied in the hydrologic field such as regression-based regional frequency analysis (RFA). The main purpose of this study is to compare two frameworks of RFA based on the artificial neural network (ANN) models: quantile regression technique (QRT-ANN) and parameter regression technique (PRT-ANN). As an output layer of the ANN model, the QRT-ANN predicts quantiles for various return periods whereas the PRT-ANN provides prediction of three parameters for the generalized extreme value distribution. Rainfall gauging sites where record length is more than 20 years were selected and their annual maximum rainfalls and various hydro-meteorological variables were used as an input layer of the ANN model. While employing the ANN model, 70% and 30% of gauging sites were used as training set and testing set, respectively. For each technique, ANN model structure such as number of hidden layers and nodes was determined by a leave-one-out validation with calculating root mean square error (RMSE). To assess the performances of two frameworks, RMSEs of quantile predicted by the QRT-ANN are compared to those of the PRT-ANN.

• 한국분말재료학회지
• /
• 제31권2호
• /
• pp.137-145
• /
• 2024

In order to predict the process window of laser powder bed fusion (LPBF) for printing metallic components, the calculation of volumetric energy density (VED) has been widely calculated for controlling process parameters. However, because it is assumed that the process parameters contribute equally to heat input, the VED still has limitation for predicting the process window of LPBF-processed materials. In this study, an explainable machine learning (xML) approach was adopted to predict and understand the contribution of each process parameter to defect evolution in Ti alloys in the LPBF process. Various ML models were trained, and the Shapley additive explanation method was adopted to quantify the importance of each process parameter. This study can offer effective guidelines for fine-tuning process parameters to fabricate high-quality products using LPBF.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1995년도 Proceedings of the Korea Automation Control Conference, 10th (KACC); Seoul, Korea; 23-25 Oct. 1995
• /
• pp.123-126
• /
• 1995

Characteristics of control system design using Universal Learning Network (U.L.N.) are that a system to be controlled and a controller are both constructed by U.L.N. and that the controller is best tuned through learning. U.L.N has the same generalization ability as N.N.. So the controller constructed by U.L.N. is able to control the system in a favorable way under the condition different from the condition of the control system in learning stage. But stability can not be realized sufficiently. In this paper, we propose a robust control method using U.L.N. and second order derivatives of U.L.N.. The proposed method can realize better performance and robustness than the commonly used Neural Network. Robust control considered here is defined as follows. Even though initial values of node outputs change from those in learning, the control system is able to reduce its influence to other node outputs and can control the system in a preferable way as in the case of no variation. In order to realize such robust control, a new term concerning the variation is added to a usual criterion function. And parameter variables are adjusted so as to minimize the above mentioned criterion function using the second order derivatives of criterion function with respect to the parameters. Finally it is shown that the controller constricted by the proposed method works in an effective way through a simulation study of a nonlinear crane system.

• 한국산업융합학회 논문집
• /
• 제18권4호
• /
• pp.259-266
• /
• 2015

This study presents new scheme for various walking pattern of biped robot under the limitted enviroments. We show that the neural network is significantly more attractive intelligent controller design than previous traditional forms of control systems. A multilayer backpropagation neural network identification is simulated to obtain a learning control solution of biped robot. Once the neural network has learned, the other neural network control is designed for various trajectory tracking control with same learning-base. The main advantage of our scheme is that we do not require any knowledge about the system dynamic and nonlinear characteristic, and can therefore treat the robot as a black box. It is also shown that the neural network is a powerful control theory for various trajectory tracking control of biped robot with same learning-vase. That is, we do net change the control parameter for various trajectory tracking control. Simulation and experimental result show that the neural network is practically feasible and realizable for iterative learning control of biped robot.

• 산업경영시스템학회지
• /
• 제43권2호
• /
• pp.120-125
• /
• 2020

Supply chain managers seek to achieve global optimization by solving problems in the supply chain's business process. However, companies in the supply chain hide the adverse information and inform only the beneficial information, so the information is distorted and cannot be the information that describes the entire supply chain. In this case, supply chain managers can directly collect and analyze supply chain activity data to find and manage the companies described by the data. Therefore, this study proposes a method to collect the order-inventory information from each company in the supply chain and detect the companies whose data characteristics are explained through deep learning. The supply chain consists of Manufacturer, Distributor, Wholesaler, Retailer, and training and testing data uses 600 weeks of time series inventory information. The purpose of the experiment is to improve the detection accuracy by adjusting the parameter values of the deep learning network, and the parameters for comparison are set by learning rate (lr = 0.001, 0.01, 0.1) and batch size (bs = 1, 5). Experimental results show that the detection accuracy is improved by adjusting the values of the parameters, but the values of the parameters depend on data and model characteristics.

• 한국통신학회논문지
• /
• 제16권1호
• /
• pp.35-45
• /
• 1991

본 논문에서는 Backpropagation 학습 이론을 사용한 다층 구조 신경 회로망을 이용하여 3차원적으로 왜곡된 항공기 인식과 항공기의 3차원 회전 방향 추정을 컴퓨터 시뮬레이션을 통하여 수행하였다. 항공기 영상으로 부터 2차원 영상에서 왜곡 불변 (distortion invariant)특정을 가지는 피치 $(L,\;{\Phi})$를 추출하여 신경 회로망 항공기 인식기의 학습(training)에 사용하였다. 그리고 신경 회로망 인식기 설계시 그 구조를 최적화 함으로써 높은 인식률을 가지는 항공기 인식기를 구성하였다. 신경 회로망 학습 과정에서 학습 이론으로는 변형된 backpropagation 학습 이론을 도입하고 아울러 학습 수행중에 학습 변수(learning parameter)값을 변화 시키는 방법을 사용하여 전체 학습 시간을 효과적으로 단축시킬 수 있었다.

• 인터넷정보학회논문지
• /
• 제22권3호
• /
• pp.9-16
• /
• 2021

This study aims to develop a human activity recognition (HAR) system as a Deep-Learning (DL) classification model, distinguishing various human activities. We solely rely on the signals from a wristband accelerometer worn by a person for the user's convenience. 3-axis sequential acceleration signal data are gathered within a predefined time-window-slice, and they are used as input to the classification system. We are particularly interested in developing a Deep-Learning model that can outperform conventional machine learning classification performance. A total of 13 activities based on the laboratory experiments' data are used for the initial performance comparison. We have improved classification performance using the Convolutional Neural Network (CNN) combined with an auto-encoder feature reduction and parameter tuning. With various publically available HAR datasets, we could also achieve significant improvement in HAR classification. Our CNN model is also compared against Recurrent-Neural-Network(RNN) with Long Short-Term Memory(LSTM) to demonstrate its superiority. Noticeably, our model could distinguish both general activities and near-identical activities such as sitting down on the chair and floor, with almost perfect classification accuracy.

• 웰빙융합연구
• /
• 제6권2호
• /
• pp.27-31
• /
• 2023

Purpose: Establishment of a real-time monitoring system for odor control in traditional markets in Gangwon-do and a system for linking prevention facilities. Research design, data and methodology: Build server and system logic based on data through real-time monitoring device (sensor-based). A temporary data generation program for deep learning is developed to develop a model for odor data. Results: A REST API was developed for using the model prediction service, and a test was performed to find an algorithm with high prediction probability and parameter values optimized for learning. In the deep learning algorithm for AI modeling development, Pandas was used for data analysis and processing, and TensorFlow V2 (keras) was used as the deep learning library. The activation function was swish, the performance of the model was optimized for Adam, the performance was measured with MSE, the model method was Functional API, and the model storage format was Sequential API (LSTM)/HDF5. Conclusions: The developed system has the potential to effectively monitor and manage odors in traditional markets. By utilizing real-time data, the system can provide timely alerts and facilitate preventive measures to control and mitigate odors. The AI modeling component enhances the system's predictive capabilities, allowing for proactive odor management.

• International Journal of Computer Science & Network Security
• /
• 제23권10호
• /
• pp.115-128
• /
• 2023

Deep learning has been incorporating various optimization techniques motivated by new pragmatic optimizing algorithm advancements and their usage has a central role in Machine learning. In recent past, new avatars of various optimizers are being put into practice and their suitability and applicability has been reported on various domains. The resurgence of novelty starts from Stochastic Gradient Descent to convex and non-convex and derivative-free approaches. In the contemporary of these horizons of optimizers, choosing a best-fit or appropriate optimizer is an important consideration in deep learning theme as these working-horse engines determines the final performance predicted by the model. Moreover with increasing number of deep layers tantamount higher complexity with hyper-parameter tuning and consequently need to delve for a befitting optimizer. We empirically examine most popular and widely used optimizers on various data sets and networks-like MNIST and GAN plus others. The pragmatic comparison focuses on their similarities, differences and possibilities of their suitability for a given application. Additionally, the recent optimizer variants are highlighted with their subtlety. The article emphasizes on their critical role and pinpoints buttress options while choosing among them.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2002년도 하계학술대회 논문집 B
• /
• pp.1165-1168
• /
• 2002

This paper presents a neural network based self-tuning fuzzy PID control scheme with variable learning rate for sensorless vector controlled induction motor drives. MRAS(Model Reference Adaptive System) is used for rotor speed estimation. When induction motor is continuously used long time. its electrical and mechanical parameters will change, which degrade the performance of PID controller considerably. This paper re-analyzes the fuzzy controller as conventional PID controller structure, introduces a single neuron with a back-propagation learning algorithm to tune the control parameters, and proposes a variable learning rate to improve the control performance. The proposed scheme is simple in structure and computational burden is small. The simulation using Matlab/Simulink and the experiment using DS1102 board show the robustness of the proposed controller to parameter variations.