• Journal of Internet Computing and Services
• /
• v.24 no.1
• /
• pp.61-69
• /
• 2023

Recently, as the fishery resources are depleted, expectations for productivity improvement by 'rearing fishery' in land farms are greatly rising. In the case of land farms, unlike ocean environments, it is easy to control and manage environmental and breeding factors, and has the advantage of being able to adjust production according to the production plan. On the other hand, unlike in the natural environment, there is a disadvantage in that operation costs may significantly increase due to the artificial management for fish growth. Therefore, profit maximization can be pursued by efficiently operating the farm in accordance with the planned target shipment. In order to operate such an efficient farm and nurture fish, an accurate growth prediction model according to the target fish species is absolutely required. Most of the growth prediction models are mainly numerical results based on statistical analysis using farm data. In this paper, we present a growth prediction model from a stochastic point of view to overcome the difficulties in securing data and the difficulty in providing quantitative expected values for inaccuracies that existing growth prediction models from a statistical point of view may have. For a stochastic approach, modeling is performed by introducing a Gaussian process regression method based on water temperature, which is the most important factor in positive growth. From the corresponding results, it is expected that it will be able to provide reference values for more efficient farm operation by simultaneously providing the average value of the predicted growth value at a specific point in time and the confidence interval for that value.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.34 no.6
• /
• pp.315-324
• /
• 2022

The experimental data obtained in a wave flume were analyzed using machine learning techniques to establish a model that predicts the input wave height of the wavemaker based on the waves that have experienced wave shoaling and to verify the performance of the established model. For this purpose, artificial neural network (NN), the most representative machine learning technique, and Gaussian process regression (GPR), one of the non-parametric regression analysis methods, were applied respectively. Then, the predictive performance of the two models was compared. The analysis was performed independently for the case of using all the data at once and for the case by classifying the data with a criterion related to the occurrence of wave breaking. When the data were not classified, the error between the input wave height at the wavemaker and the measured value was relatively large for both the NN and GPR models. On the other hand, if the data were divided into non-breaking and breaking conditions, the accuracy of predicting the input wave height was greatly improved. Among the two models, the overall performance of the GPR model was better than that of the NN model.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.53 no.7
• /
• pp.536-543
• /
• 2004

In this paper, we proposed a new method of modeling a neuro-fuzzy system using a hybrid clustering algorithm. The initial parameters and the number of clusters of the proposed system are optimally chosen simultaneously with respect to the process of regression, which is a unique characteristics of the proposed system. The proposed algorithm presented in this work improves the overall performance of the proposed a neuro-fuzzy system by choosing a proper number of clusters adaptively according the characteristics of given data. The process of clustering is performed by deciding on the number of classes, which yields the property of convergence of the system. In experiments, the superiority of the proposed neuro-fuzzy system is demonstrated, especially the process of optimizing parameters and clustering of learning speed.

• Journal of the Society of Naval Architects of Korea
• /
• v.49 no.4
• /
• pp.359-366
• /
• 2012

This is the final one of the two companion papers dealing with accuracy of accumulated fatigue damage estimation under wide band process. It is stated that four kinds of wide band models exist: typed of equivalent stress, combined PDF, correction factor, and damage combination. For the idealized ESDs from full scale measurement data on an 8100TEU container vessel, fatigue damages are compared for a narrow band prediction model based on Rayleigh PDF and five wide band fatigue prediction models of Dirlik, Wirsching-Light, Jiao-Moan, Benasciutti and DNV. DNV model consistently overestimates fatigue damages regardless of variation of ESDs. Predictions by Jiao-Moan model, which is understood as standard method for design of offshore platforms, are also in conservative side. Best accuracy is found from the results by Dirlik and Benasciutti models, but Benasciutti model is preferred since it can easily combined with narrow band fatigue damage based on Rayleigh PDF.

• Tunnel and Underground Space
• /
• v.32 no.6
• /
• pp.586-597
• /
• 2022

In this study, a model was developed for the estimating the locational thickness information of the upper clay layer to be used for the consolidation vulnerability evaluation in the Nakdong river estuary. To estimate ground layer thickness information, we developed four spatial estimation models using machine learning algorithms, which are RF (Random Forest), SVR (Support Vector Regression) and GPR (Gaussian Process Regression), and geostatistical technique such as Ordinary Kriging. Among the 4,712 borehole data in the study area collected for model development, 2,948 borehole data with an upper clay layer were used, and Pearson correlation coefficient and mean squared error were used to quantitatively evaluate the performance of the developed models. In addition, for qualitative evaluation, each model was used throughout the study area to estimate the information of the upper clay layer, and the thickness distribution characteristics of it were compared with each other.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.24 no.2
• /
• pp.280-291
• /
• 1987

In this paper, the linear autoregressive model of EMG signal for four basic arm functions was presented and parameters for each function were estimated. The signal identification was carried out using function discrimination algorithm. It was validated that EMG signal was a widesense stationary process and the linear autoregressive model of EMG signal was constructed through approximating it to Gaussian process. It was confined that Levinson-Durbin algoridthm is a more appropriate one than the recursive least square method for parameter estimation of the linear model. Optimal function discrimination was acquired when sampling frequency was 500Hz and two electrodes were attached to bicep and tricep muscle, respectively. Parameter values were independent of variance and the number of minimum data for function discrimination was 200. Bayesian discrimination method turned out to be a better one than parallel filtering method for functional discrimination recognition.

• Communications for Statistical Applications and Methods
• /
• v.21 no.5
• /
• pp.423-433
• /
• 2014

Test statistics using cumulative sums of residuals have been widely used in various regression models including generalized linear models(GLM). Recently, Pan and Lin (2005) extended this testing procedure to the generalized linear mixed models(GLMM) having random effects, in which we encounter difficulties in computing the marginal likelihood that is expressed as an integral of random effects distribution. The Gaussian quadrature algorithm is commonly used to approximate the marginal likelihood. Many commercial statistical packages provide an option to apply this type of goodness-of-fit test in GLMs but available programs are very rare for GLMMs. We suggest a computational algorithm to implement the testing procedure in GLMMs by a freely accessible R package, and also illustrate through practical examples.

• Structural Engineering and Mechanics
• /
• v.32 no.1
• /
• pp.179-191
• /
• 2009

A stochastic response spectrum method is proposed for simple evaluation of the structural response of an actively controlled aseismic structure. The response spectrum is constructed assuming a linear structure with an active mass damper (AMD) system, and an earthquake wave model given by the product of a non-stationary envelope function and a stationary Gaussian random process with Kanai-Tajimi power spectral density. The control design is executed using a linear quadratic Gaussian control strategy for an enlarged state space system, and the response amplification factor is given by the combination of the obtained statistical response values and extreme value theory. The response spectrum thus produced can be used for simple dynamical analyses. The response factors obtained by this method for a multi-degree-of-freedom structure are shown to be comparable with those determined by numerical simulations, demonstrating the validity and utility of the proposed technique as a simple design tool. This method is expected to be useful for engineers in the initial design stage for structures with active aseismic control.

• Journal of the Korean Statistical Society
• /
• v.26 no.2
• /
• pp.231-243
• /
• 1997

In this paper we investigate weak convergence of the intergral processes whose index set is the non-compact infinite time interval. Our first goal is to develop the empirical central limit theorem as random elements of [0, .infty.) for an integral process which is constructed from iid variables. In developing the weak convergence as random elements of D[0, .infty.), we will use a result of Ossiander(4) whose proof heavily depends on the total boundedness of the index set. Our next goal is to establish the empirical central limit theorem for the Kaplan-Meier integral process as random elements of D[0, .infty.). In achieving the the goal, we will use the above iid result, a representation of State(6) on the Kaplan-Meier integral, and a lemma on the uniform order of convergence. The first result, in some sense, generalizes the result of empirical central limit therem of Pollard(5) where the process is regarded as random elements of D[-.infty., .infty.] and the sample paths of limiting Gaussian process may jump. The second result generalizes the first result to random censorship model. The later also generalizes one dimensional central limit theorem of Stute(6) to a process version. These results may be used in the nonparametric statistical inference.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.38 no.6
• /
• pp.401-415
• /
• 1989

A simple but accurate analytical model for the lateral channel electric field in gate-offset structured Lightly Doped Drain MOSFET has been developed. Our model assumes Gaussian doping profile, rather than simple uniform doping, for the lightly doped region and our model can be applied to LDD structures where the junction depth of LDD is not identical to the heavily doped drain. The validity of our model has been proved by comparing our analytical results with two dimensional device simulations. Due to its simplicity, our model gives a better understanding of the mechanisms involved in reducing the electric field in the LDD MOSFET. The model shows clearly the dependencies of the lateral channel electric field on the drain and gate bias conditions and process, design parameters. Advantages of our analytical model over costly 2-D device simulations is to identify the effects of various parameters, such as oxide thickness, junction depth, gate/drain bias, the length and doping concentration of the lightly doped region, on the peak electric field that causes hot-electron pohenomena, individually. Our model can also find the optimum doping concentration of LDD which minimizes the peak electric field and hot-electron effects.