• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2007.05a
• /
• pp.389-398
• /
• 2007

Case-based reasoning (CBR) is one of the most popular prediction techniques for medical diagnosis because it is easy to apply, has no possibility of overfitting, and provides a good explanation for the output. However, it has a critical limitation - its prediction performance is generally lower than other artificial intelligence techniques like artificial neural networks (ANNs). In order to obtain accurate results from CBR, effective retrieval and matching of useful prior cases for the problem is essential, but it is still a controversial issue to design a good matching and retrieval mechanism for CBR systems. In this study, we propose a novel approach to enhance the prediction performance of CBR. Our suggestion is the simultaneous optimization of feature weights, instance selection, and the number of neighbors that combine using genetic algorithms (GAs). Our model improves the prediction performance in three ways - (1) measuring similarity between cases more accurately by considering relative importance of each feature, (2) eliminating redundant or erroneous reference cases, and (3) combining several similar cases represent significant patterns. To validate the usefulness of our model, this study applied it to a real-world case for evaluating cytological features derived directly from a digital scan of breast fine needle aspirate (FNA) slides. Experimental results showed that the prediction accuracy of conventional CBR may be improved significantly by using our model. We also found that our proposed model outperformed all the other optimized models for CBR using GA.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.14 no.3
• /
• pp.635-644
• /
• 1994

A mathematical model was developed to predict the concentrations of various nutrients in supernatants during aerobic digestion which is suitable to be employed in small wastewater treatment plants with such advantages as low capital cost and stable process. Significant reactions were determined with observing the behavior of nitrgen and phosphorus, and the model equations were built up in the form of simultaneous differential equations considering Mass Balance. Laboratory batch experiments were carried out at $20^{\circ}C$ and pH $7.5{\pm}0.5$ on the aerobic digestion of waste activated sludge at different solid levels. Nonlinear regression analysis was performed to estimate various reaction rate constants. The developed model can predict the behavior of Biomass N, dissolved organic N, $NH_4{^-}$-N, $NO_x{^-}$-N, and Biomass P, dissolved organic P, $PO_4{^-}$-P in aerobic digestion process. In this study, the results of simulation showed that dissolved nutrients had more effects on supernatants than nutrients in biomass, and phosphorus was more effective on supernatants than nitrogen.

• Bulletin of the Korean Space Science Society
• /
• 2008.10a
• /
• pp.27.3-28
• /
• 2008

In this study, we report a new Monte Carlo ray tracing technique for estimating GOCI (Geostationary Ocean Color Instrument) radiative transfer characteristics and imaging performance simultaneously. First, a full scale GOCI optical model was constructed with measured characteristics at the component level and placed in the geostationary orbit. An optical model of approximated GOCI target area centered at the Korean penninsular was then built using the USGS coastal line data and representative land and sea surface reflectivity data. The light rays launched from a simulated sun model travel to the Earth surface, where they are reflected and scattered. Some of the light rays that are headed to the GOCI model in the orbit were selected and traced, as they have entered into the GOCI aperture. As they pass through each GOCI optical part, the ray path and intensity are adjusted according to the measured characteristics for reflection, transmission, refractive index and surface scattering. The ray-traced imaging and radiative transfer performance indicators confirm that the computer generated GOCI optical system with measured characteristics can be used for in-orbit operation simulation following the designed measurement sequence. The computational technique and its implications as a operation support tool are discussed.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2006.11a
• /
• pp.599-617
• /
• 2006

A network model and a Genetic Algorithm(GA) is proposed to solve the simultaneous estimation of the trip distribution and traffic assignment from traffic counts in the congested networks in a logit-based Stochastic User Equilibrium (SUE). The model is formulated as a problem of minimizing the non-linear objective functions with the linear constraints. In the model, the flow-conservation constraints of the network are utilized to restrict the solution space and to force the link flows meet the traffic counts. The objective of the model is to minimize the discrepancies between the link flows satisfying the constraints of flow-conservation, trip production from origin, trip attraction to destination and traffic counts at observed links and the link flows estimated through the traffic assignment using the path flow estimator in the legit-based SUE. In the proposed GA, a chromosome is defined as a vector representing a set of Origin-Destination Matrix (ODM), link flows and travel-cost coefficient. Each chromosome is evaluated from the corresponding discrepancy, and the population of the chromosome is evolved by the concurrent simplex crossover and random mutation. To maintain the feasibility of solutions, a bounded vector shipment is applied during the crossover and mutation.

• Journal of Communications and Networks
• /
• v.5 no.2
• /
• pp.157-166
• /
• 2003

This paper deals with the problem of channel identification for Single Input Multiple Output (SIMO) slow fading channels using clustering algorithms. Due to the intrinsic memory of the discrete-time model of the channel, over short observation periods, the received data vectors of the SIMO model are spread in clusters because of the AWGN noise. Each cluster is practically centered around the ideal channel output labels without noise and the noisy received vectors are distributed according to a multivariate Gaussian distribution. Starting from the Markov SIMO channel model, simultaneous maximum ikelihood estimation of the input vector and the channel coefficients reduce to one of obtaining the values of this pair that minimizes the sum of the Euclidean norms between the received and the estimated output vectors. Viterbi algorithm can be used for this purpose provided the trellis diagram of the Markov model can be labeled with the noiseless channel outputs. The problem of identification of the ideal channel outputs, which is the focus of this paper, is then equivalent to designing a Vector Quantizer (VQ) from a training set corresponding to the observed noisy channel outputs. The Linde-Buzo-Gray (LBG)-type clustering algorithms [1] could be used to obtain the noiseless channel output labels from the noisy received vectors. One problem with the use of such algorithms for blind time-varying channel identification is the codebook initialization. This paper looks at two critical issues with regards to the use of VQ for channel identification. The first has to deal with the applicability of this technique in general; we present theoretical results for the conditions under which the technique may be applicable. The second aims at overcoming the codebook initialization problem by proposing a novel approach which attempts to make the first phase of the channel estimation faster than the classical codebook initialization methods. Sample simulation results are provided confirming the effectiveness of the proposed initialization technique.

• Safety and Health at Work
• /
• v.11 no.3
• /
• pp.322-334
• /
• 2020

Background: Spontaneous combustion of coal is one of the factors which causes direct or indirect gas and dust explosion, mine fire, the release of toxic gases, loss of reserve, and loss of miners' life. To avoid these incidents, the prediction of spontaneous combustion is essential. The safety of miner's in the mining field can be assured if the prediction of a coal fire is carried out at an early stage. Method: Adularya Underground Coal Mine which is fully mechanized with longwall mining method was selected as a case study area. The data collected for 2017, by sensors from ten gas monitoring stations were used for the simulation and prediction of a coal fire. In this study, the fuzzy logic model is used because of the uncertainties, nonlinearity, and imprecise variables in the data. For coal fire prediction, CO, O₂, N₂, and temperature were used as input variables whereas fire intensity was considered as the output variable.The simulation of the model is carried out using the Mamdani inference system and run by the Fuzzy Logic Toolbox in MATLAB. Results: The results showed that the fuzzy logic system is more reliable in predicting fire intensity with respect to uncertainties and nonlinearities of the data. It also indicates that the 1409 and 610/2B gas station points have a greater chance of causing spontaneous combustion and therefore require a precautional measure. Conclusion: The fuzzy logic model shows higher probability in predicting fire intensity with the simultaneous application of many variables compared with Graham's index.

• The Korean Journal of Applied Statistics
• /
• v.29 no.5
• /
• pp.807-822
• /
• 2016

The composite lognormal-GPD models (LN-GPD) enjoys both merits from log-normality for the body of distribution and GPD for the thick tailedness of the observation. However, in the estimation perspective, LN-GPD model performs poorly due to numerical instability. Therefore, a two-stage procedure, that estimates threshold first then estimates other parameters later, is a natural method to consider. This paper considers five nonparametric threshold estimation methods widely used in extreme value theory and compares their performance in LN-GPD parameter estimation. A simulation study reveals that simultaneous maximum likelihood estimation performs good in threshold estimation, but very poor in tail index estimation. However, the nonparametric method performs good in tail index estimation, but introduced bias in threshold estimation. Our method is illustrated to the service time of an Israel bank call center and shows that the LN-GPD model fits better than LN or GPD model alone.

• Korean Journal of Environmental Biology
• /
• v.20 no.4
• /
• pp.277-286
• /
• 2002

An increase in the overall biological effect under the combined action of ionizing radiation with another inactivating agent can be explained in two ways. One is the supposition that synergism may attribute to a reduced cellular capacity of damn-ge repair after the combined action. The other is the hypothesis that synergism may be related to an additional lethal or potentially lethal damage that arises from the interaction of sublesions induced by both agents. These sublesions ave considered to be in-effective when each agent is applied separately. Based on this hypothesis, a simple mathematical model was established. The model can predict the greatest value of the synergistic effect, and the dependence of synergy on the intensity of agents applied, as well. This paper deals with the model validation and the peculiarity of simultaneous action of various factors with radiation on biological systems such as bacteriophage, bacterial spores, yeast and mammalian cells. The common rules of the synergism aye as follows. (1) For any constant rate of exposure, the synergy can be observed only within a certain temperature range. The temperature range which synergistically increases the effects of radiation is shifted to the lower temperature fer thermosensitive objects. Inside this range, there is a specific temperature that maximizes the synergistic effect. (2) A decrease in the exposure rate results in a decrease of this specific temperature to achieve the greatest synergy and vice versa. For a constant temperature at which the irradiation occurs, synergy can be observed within a certain dose rate range. Inside this range an optimal intensity of the physical agent may be indicated, which maximizes the synergy. As the exposure temperature reduces, the optimal intensity decreases and vice versa. (3) The recovery rate after combined action is decelerated due to an increased number of irreversible damages. The probability of recovery is independent of the exposure temperature for yeast cells irradiated with ionizing or UV radiation. Chemical inhibitors of cell recovery act through the formation of irreversible damage but not via damaging the recovery process itself.

• Journal of Biosystems Engineering
• /
• v.35 no.6
• /
• pp.401-407
• /
• 2010

In this study, computer simulations were conducted to assess the use of a circulating concurrent-flow dryer for rapeseed drying and to determined the effect of this drying method on the germination ratio of rapeseed after the drying process was complete. The simultaneous heat and mass transfer between air and rapeseed in a concurrent-flow dryer was examined by simulation. The drying simulation was based on several parameters with sequent time series. Equations concerning air psychrometrics, physical properties, thermal properties, equilibrium moisture content, thin layer drying of rapeseed, etc. were all combined to solve the simulation models. Based on energy and mass transfer in the concurrent-flow drying model, a simulation program for the circulating concurrent-flow rapeseed dryer was built along with a detailed description of the mathematical solution to the model. A pilot scale circulating concurrent-flow dryer(200 kg/batch) was used to verify the fitness of the simulation program. A comparison between the experimental data and the model predicted results was presented and discussed. The drying parameters and germination ratio were analyzed and the accuracy of the simulation program was evaluated. The simulation program proved to be reliable and was shown to be a convenient tool for predicting rapeseed drying and germination ratio of rapeseed in a concurrent-flow dryer.

• The Korean Journal of Applied Statistics
• /
• v.30 no.1
• /
• pp.57-68
• /
• 2017

Clustering algorithms attempt to find a partition of a finite set of objects in to a potentially predetermined number of nonempty subsets. Gibbs sampling of a normal mixture of linear mixed regressions with a Dirichlet prior distribution calculates posterior probabilities when the number of clusters was known. Our approach provides simultaneous partitioning and parameter estimation with the computation of classification probabilities. A Monte Carlo study of curve estimation results showed that the model was useful for function estimation. Examples are given to show how these models perform on real data.