• KIISE Transactions on Computing Practices
• /
• v.23 no.7
• /
• pp.440-445
• /
• 2017

Due to the advance of e-commerce systems, the number of people using online shopping and products has significantly increased. Therefore, the need for an accurate recommendation system is becoming increasingly more important. Recurrent neural network is a deep-learning algorithm that utilizes sequential information in training. In this paper, an evaluation is performed on the application of recurrent neural networks to recommendation systems. We evaluated three recurrent algorithms (RNN, LSTM and GRU) and three optimal algorithms(Adagrad, RMSProp and Adam) which are commonly used. In the experiments, we used the TensorFlow open source library produced by Google and e-commerce session data from RecSys Challenge 2015. The results using the optimal hyperparameters found in this study are compared with those of RecSys Challenge 2015 participants.

• Journal of Ocean Engineering and Technology
• /
• v.29 no.5
• /
• pp.366-372
• /
• 2015

Ocean energy is one of the most promising renewable energy resources. In particular, South Korea is one of the countries where it is economically and technically feasible to develop tidal current power plants to use tidal current energy. In this study, based on the design code for HARP_Opt (Horizontal axis rotor performance optimizer) developed by NREL (National Renewable Energy Laboratory) in the United States, and applying the BEMT (Blade element momentum theory) and GA (Genetic algorithm), the optimal shape design and performance evaluation of the horizontal axis rotor for a 200-kW-class tidal current turbine were performed using different numbers of blades (two or three) and a pitch control method (variable pitch or fixed pitch). As a result, the VSFP (Variable Speed Fixed Pitch) turbine with three blades showed the best performance. However, the performances of four different cases did not show significant differences. Hence, it is necessary when selecting the final design to consider the structural integrity related to the fatigue, along with the economic feasibility of manufacturing the blades.

• Computers and Concrete
• /
• v.32 no.2
• /
• pp.217-232
• /
• 2023

Numerous studies have been performed on the behavior of pile foundations in cold regions. This study first attempted to employ artificial neural networks (ANN) to predict pile-bearing capacity focusing on pile data recorded primarily on cold regions. As the ANN technique has disadvantages such as finding global minima or slower convergence rates, this study in the second phase deals with the development of an ANN-based predictive model improved with an Elephant herding optimizer (EHO), Dragonfly Algorithm (DA), Genetic Algorithm (GA), and Evolution Strategy (ES) methods for predicting the piles' bearing capacity. The network inputs included the pile geometrical features, pile area (m²), pile length (m), internal friction angle along the pile body and pile tip (Ø°), and effective vertical stress. The MLP model pile's output was the ultimate bearing capacity. A sensitivity analysis was performed to determine the optimum parameters to select the best predictive model. A trial-and-error technique was also used to find the optimum network architecture and the number of hidden nodes. According to the results, there is a good consistency between the pile-bearing DA-MLP-predicted capacities and the measured bearing capacities. Based on the R2 and determination coefficient as 0.90364 and 0.8643 for testing and training datasets, respectively, it is suggested that the DA-MLP model can be effectively implemented with higher reliability, efficiency, and practicability to predict the bearing capacity of piles.

• Geomechanics and Engineering
• /
• v.31 no.4
• /
• pp.339-352
• /
• 2022

Studying slope stability is an important branch of civil engineering. In this way, engineers have employed machine learning models, due to their high efficiency in complex calculations. This paper examines the robustness of various novel optimization schemes, namely equilibrium optimizer (EO), Harris hawks optimization (HHO), water cycle algorithm (WCA), biogeography-based optimization (BBO), dragonfly algorithm (DA), grey wolf optimization (GWO), and teaching learning-based optimization (TLBO) for enhancing the performance of adaptive neuro-fuzzy inference system (ANFIS) in slope stability prediction. The hybrid models estimate the factor of safety (F_S) of a cohesive soil-footing system. The role of these algorithms lies in finding the optimal parameters of the membership function in the fuzzy system. By examining the convergence proceeding of the proposed hybrids, the best population sizes are selected, and the corresponding results are compared to the typical ANFIS. Accuracy assessments via root mean square error, mean absolute error, mean absolute percentage error, and Pearson correlation coefficient showed that all models can reliably understand and reproduce the F_S behavior. Moreover, applying the WCA, EO, GWO, and TLBO resulted in reducing both learning and prediction error of the ANFIS. Also, an efficiency comparison demonstrated the WCA-ANFIS as the most accurate hybrid, while the GWO-ANFIS was the fastest promising model. Overall, the findings of this research professed the suitability of improved intelligent models for practical slope stability evaluations.

• KIPS Transactions on Computer and Communication Systems
• /
• v.10 no.3
• /
• pp.71-80
• /
• 2021

The selection of an appropriate neural network algorithm is an important step for accurate data prediction in machine learning. Many algorithms based on basic artificial neural networks have been devised to efficiently predict future data. These networks include deep neural networks (DNNs), recurrent neural networks (RNNs), long short-term memory (LSTM) networks, and gated recurrent unit (GRU) neural networks. Developers face difficulties when choosing among these networks because sufficient information on their performance is unavailable. To alleviate this difficulty, we evaluated the performance of each algorithm by comparing their errors and processing times. Each neural network model was trained using a tax dataset, and the trained model was used for data prediction to compare accuracies among the various algorithms. Furthermore, the effects of activation functions and various optimizers on the performance of the models were analyzed The experimental results show that the GRU and LSTM algorithms yields the lowest prediction error with an average RMSE of 0.12 and an average R2 score of 0.78 and 0.75 respectively, and the basic DNN model achieves the lowest processing time but highest average RMSE of 0.163. Furthermore, the Adam optimizer yields the best performance (with DNN, GRU, and LSTM) in terms of error and the worst performance in terms of processing time. The findings of this study are thus expected to be useful for scientists and developers.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.32 no.10
• /
• pp.12-19
• /
• 2004

An aerodynamic design optimization system for three-dimensional wing was developed as a part of the future MDO framework. The present design optimization system includes four modules such as geometry design, grid generation, flow solver and optimizer. All modules were based on commercial softwares and programmed to have automated execution capability in batch mode utilizing built-in script and journaling. The integration of all modules into the system was accomplished through programming using Visual Basic language. The distributed computational environment based on network communication was established to save computational time especially for time-consuming aerodynamic analyses. The distributed aerodynamic computations were performed in conjunction with the global optimization algorithm of response surface method, instead of using usual parallel computation based on domain decomposition. The application of the design system in the drag minimization problem demonstrated considerably enhanced efficiency of the design process while the final design showed reasonable results of reduced drag.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.26 no.6
• /
• pp.108-114
• /
• 1989

Introducing and optimizing bit-significance to the Hopfield model, ten highly correlated binary images, i.e., numbers "0" to "9", are successfully stored and retrieved in a $6{}8$ node system. Unlike many other neural network models, this model has stronger error correction capability for correlated images such as "6","8","3", and "9". The bit significance optimization is regarded as an adaptive learning process based on least-mean-square error algorithm, and may be implemented with Widrow-Hoff neural nets optimizer. A design for electro-optic implementation including the adaptive optimization networks is also introduced.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.16 no.4
• /
• pp.630-648
• /
• 1992

The minimum weight design for simply-supported isotropic or symmetrically laminated stiffened cylindrical shells subjected to various loads (axial compression or combined loads) is studied by a nonlinear mathematical search algorithm. The minimum weight design in accomplished with the CONMIN optimizer by Vanderplaats. Several types of buckling modes with maximum allowable stresses and strains are included as constraints in the minimum weight design process, such as general buckling, panel buckling with either stingers or rings smeared out, local skin buckling, local crippling of stiffener segments, and general, panel and local skin buckling including stiffener rolling. The approach allows the consideration of various shapes of stiffening members. Rectangular, I, or T type stringers and rectangular rings are used for stiffened cylindrical shells. Several design examples are analyzed and compared with those in the previous literatures. The unstiffened glass/epoxy, graphite/epoxy(T300/5208), and graphite/epoxy aluminum honeycomb cylindrical shells and stiffened graphite/epoxy cyindrical shells under axial compression are analyzed through the present approach.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.39 no.2
• /
• pp.28-36
• /
• 2016

Recently, owing to the development of ICT industry and wide spread of smart phone, the number of people who use car sharing service are increased rapidly. Currently two-way car sharing system with same rental and return locations are mainly operated since this system can be easily implemented and maintained. Currently the demand of one-way car sharing service has increase explosively. But this system have several obstacle in operation, especially, vehicle stock imbalance issues which invoke vehicle relocation. Hence in this study, we present an optimization approach to depot location and relocation policy in one-way car sharing systems. At first, we modelled as mixed-integer programming models whose objective is to maximize the profits of a car sharing organization considering all the revenues and costs involved and several constraints of relocation policy. And to solve this problem efficiently, we proposed a new method based on particle swarm optimization, which is one of powerful meta-heuristic method. The practical usefulness of the approach is illustrated with a case study involving satellite cities in Seoul Metrolitan Area including several candidate area where this kind systems have not been installed yet and already operating area. Our proposed approach produced plausible solutions with rapid computational time and a little deviation from optimal solution obtained by CPLEX Optimizer. Also we can find that particle swarm optimization method can be used as efficient method with various constraints. Hence based on this results, we can grasp a clear insight into the impact of depot location and relocation policy schemes on the profitability of such systems.

• Industrial Engineering and Management Systems
• /
• v.4 no.1
• /
• pp.54-67
• /
• 2005

This paper provides an evaluation of an optimization-based, multiple-input double-output (MIDO) run-to-run (R2R) control scheme for general semiconductor manufacturing processes. The controller in this research, termed adaptive dual response optimizing controller (ADROC), can serve as a process optimizer as well as a recipe regulator between consecutive runs of wafer fabrication. In evaluation, it is assumed that the equipment model could be appropriately described by a pair of second-order polynomial functions in terms of a set of controllable variables. Of practical relevance is to consider a drifting effect in the equipment model since in common semiconductor practice the process tends to drift due to machine aging and tool wearing. We select a typical application of R2R control to chemical mechanical planarization (CMP) in semiconductor manufacturing in this evaluation, and there are five different CMP process scenarios demonstrated, including mean shift, variance increase, and IMA disturbances. For the controller, ADROC, an on-line estimation technique is implemented in a self-tuning (ST) control manner for the adaptation purpose. Subsequently, an ad hoc global optimization algorithm based on the dual response approach, arising from the response surface methodology (RSM) literature, is used to seek the optimum recipe within the acceptability region for the execution of next run. The main components of ADROC are described and its control performance is assessed. It reveals from the evaluation that ADROC can provide excellent control actions for the MIDO R2R situations even though the process exhibits complicated, nonlinear interaction effects between control variables, and the drifting disturbances.