• Proceedings of the Korean Vacuum Society Conference
• /
• 2012.02a
• /
• pp.241-241
• /
• 2012

Semiconductor industry has been taking the advantage of improvements in process technology in order to maintain reduced device geometries and stringent performance specifications. This results in semiconductor manufacturing processes became hundreds in sequence, it is continuously expected to be increased. This may in turn reduce the yield. With a large amount of investment at stake, this motivates tighter process control and fault diagnosis. The continuous improvement in semiconductor industry demands advancements in process control and monitoring to the same degree. Any fault in the process must be detected and classified with a high degree of precision, and it is desired to be diagnosed if possible. The detected abnormality in the system is then classified to locate the source of the variation. The performance of a fault detection system is directly reflected in the yield. Therefore a highly capable fault detection system is always desirable. In this research, time series modeling of the data from an etch equipment has been investigated for the ultimate purpose of fault diagnosis. The tool data consisted of number of different parameters each being recorded at fixed time points. As the data had been collected for a number of runs, it was not synchronized due to variable delays and offsets in data acquisition system and networks. The data was then synchronized using a variant of Dynamic Time Warping (DTW) algorithm. The AutoRegressive Integrated Moving Average (ARIMA) model was then applied on the synchronized data. The ARIMA model combines both the Autoregressive model and the Moving Average model to relate the present value of the time series to its past values. As the new values of parameters are received from the equipment, the model uses them and the previous ones to provide predictions of one step ahead for each parameter. The statistical comparison of these predictions with the actual values, gives us the each parameter's probability of fault, at each time point and (once a run gets finished) for each run. This work will be extended by applying a suitable probability generating function and combining the probabilities of different parameters using Dempster-Shafer Theory (DST). DST provides a way to combine evidence that is available from different sources and gives a joint degree of belief in a hypothesis. This will give us a combined belief of fault in the process with a high precision.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.33 no.3
• /
• pp.250-256
• /
• 2000

The purpose of this study was to investigate the adsorption kinetics of heavy metals (Cu, Cd and Pb) using three tidal flat sediments and two yellow loesses. The relationship between adsorption rate calculated by non-linear regression model and chemical parameters was estimated. The contents of ignitiot loss (I.L.) am Fe, Mn and Al oxides of yellow loess were higher $1.5{\~}6 times$ than those of tidal flat sediments. But the contents of silt and clay of tidal flat sediment in Eueunri was higher than others. Heavy metals adsorption were occured rapidly in the intial 30 min and the concentration of adsorbed heavy metals were $4.1{\~}14.7\;{\mu}g/g\;for\;Cu,\;2.8{\~}16.7\;{\mu}g/g\;for\;Cd\;and\;43.4{\~}101.7\;{\mu}g/g$ for Pb, showing a high cumulative adsorption of $8{\~}70{\%}\;for\;Cu,\;18{\~}31{\%}\;for\;Cd and\;19{\~}52{\%}$ for Pb after 3hr. In initial concentration of $0.5{\times}10^(-5)M$, adsorption rate of heavy metals by the tidal flat sediments and yellow loesses was the sequence Pb>Cu^gt;Cd. The adsorption kinetics of Cu, Cd and Pb was found to be one-site kinetic model. Especially, in the case of Cu, there was a high negative ($R^2= -0.88{\~}-0.99$) linear correlation between chemical parameter such as I.L., Al oxide, silt and clay, and adsorption rate coefficients ($K_a$) calculated by non-linear model.

• Journal of The Korean Association For Science Education
• /
• v.34 no.2
• /
• pp.135-146
• /
• 2014

The purpose of this study is to explore evaluation criteria that pre-service elementary teachers employ as they evaluate and select models to explain electric circuits. Thirty junior students in a university of education have participated in the study as a part of the science education course in which they were enrolled. The lessons for the participants have been organized as a cyclic sequence of different modeling pedagogies including the expressive, experimental, and evaluative modeling. The pre-service teachers have been given five electric circuits in order and asked to create models and further develop them through peer discussion. Their modeling activities have been video- or audio-recorded, and the recordings and their transcripts have been analyzed using a framework of model evaluation criteria. It reveals that the types and frequencies of evaluation criteria used are different between situations of model development and model selection. While empirical and theoretical criteria have been used dominantly in both situations, more various criteria have been employed in the situation where the pre-service teachers selected one model among alternatives. Implications for science education and science education research have been suggested.

• The Journal of the Acoustical Society of Korea
• /
• v.19 no.3
• /
• pp.35-44
• /
• 2000

In this paper, we propose a new trajectory model for characterizing segmental features and their interaction based upon a general framework of hidden Markov models. Each segment, a sequence of vectors, is represented by a trajectory of observed sequences. This trajectory is obtained by applying a new design matrix which includes transitional information on contiguous frames, and is characterized as a polynomial regression function. To apply the trajectory to the segmental HMM, the frame features are replaced with the trajectory of a given segment. We also propose the likelihood of a given segment and the estimation of trajectory parameters. The obervation probability of a given segment is represented as the relation between the segment likelihood and the estimation error of the trajectories. The estimation error of a trajectory is considered as the weight of the likelihood of a given segment in a state. This weight represents the probability of how well the corresponding trajectory characterize the segment. The proposed model can be regarded as a generalization of a conventional HMM and a parametric trajectory model. The experimental results are reported on the TIMIT corpus and performance is show to improve significantly over that of the conventional HMM.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.1
• /
• pp.1-8
• /
• 2021

In the political field of news articles, there are polarized and biased characteristics such as conservative and liberal, which is called political bias. We constructed keyword-based dataset to classify bias of news articles. Most embedding researches represent a sentence with sequence of morphemes. In our work, we expect that the number of unknown tokens will be reduced if the sentences are constituted by subwords that are segmented by the language model. We propose a document embedding model with subword tokenization and apply this model to SVM and feedforward neural network structure to classify the political bias. As a result of comparing the performance of the document embedding model with morphological analysis, the document embedding model with subwords showed the highest accuracy at 78.22%. It was confirmed that the number of unknown tokens was reduced by subword tokenization. Using the best performance embedding model in our bias classification task, we extract the keywords based on politicians. The bias of keywords was verified by the average similarity with the vector of politicians from each political tendency.

• Composites Research
• /
• v.32 no.5
• /
• pp.216-221
• /
• 2019

Piecewise integrated composite (PIC) beam has different stacking sequences for several regions with respect to their superior load-resisting capabilities. On the interest of current research is to improve bending characteristics of PIC beam, with assigning specific stacking sequence to a specific region with the help of machine learning techniques. 240 elements of from the FE model were chosen to be reference points. Preliminary FE analysis revealed triaxialities at those regularly distributed reference points to obtain learning data creation of machine learning. Triaxiality values catagorise the type of loading i.e. tension, compression or shear. Machine learning model was formulated by learning data as well as hyperparameters and proper load fidelity was suggested by tuned values of hyperparameters, however, comparatively higher nonlinearity intensive region, such as side face of the beam showed poor load fidelity. Therefore, irregular distribution of reference points, i.e., dense reference points were distributed in the severe changes of loading, on the contrary, coarse distribution for rare changes of loading, was prepared for machine learning model. FE model with irregularly distributed reference points showed better load fidelity compared to the results from the model with regular distribution of reference points.

• Journal of Korean Elementary Science Education
• /
• v.25 no.3
• /
• pp.307-317
• /
• 2006

Those students with ability and interest in science should be supported to develop their potential and to reach high levels of achievement in science and technology. In order to ensure that gifted pupils are able to enhance their creativity as well as research abilities, appropriate learning programs and environments are essential. One of the various teaching and learning models for the gifted in science is the discovery learning model based on inductive science activities. There is a clear line of continuity between knowledge discovery at the forefront of research and student's learning activities. If students receive excellent training in organizing scientific concepts for themselves, they will be able to skillfully apply appropriate scientific concepts and solve problems when facing unfamiliar situations. It is very important to offer an appropriate learning environment to maximize the learning effect whilst, at the same time, understanding individual student's characteristics. In this study, the authors took great pains to research effective learning environments for gifted science students. Firstly, appropriate classroom learning environments thought by the teacher to offer the most potential were investigated. 3 different classes in which a revised teaching and learning environment was applied in sequence were examined. Inquiries were conducted into students' activities and achievement through observation, interviews, and examination of students' worksheets. A Science Education expert and 5 elementary school teachers specializing in gifted education also observed the class to examine the specific character of gifted science students. A number of suggestions in discovery learning classes for elementary students gifted in science are possible; 1) Readiness is essential in attitudes related to the inquiry. 2) The interaction between students should be developed. A permissive atmosphere is needed in small group activities. 3) Students require training in listening to others. In a whole class discussion, a permissive atmosphere needs to be restricted somewhat in order to promote full and inclusive discussion. 4) Students should have a chance to practice induction and abduction methods in solving problems.

• Composites Research
• /
• v.35 no.5
• /
• pp.317-321
• /
• 2022

In this study, the PIC design method with machine learning that automatically assigning different stacking sequences according to loading types was applied bumper beam. The input value and labels of the training data for applying machine learning were defined as coordinates and loading types of reference elements that are part of the total elements, respectively. In order to compare the 2D and 3D implementation method, which are methods of representing coordinate value, training data were generated, and machine learning models were trained with each method. The 2D implementation method is divided FE model into each face and generating learning data and training machine learning models accordingly. The 3D implementation method is training one machine learning model by generating training data from the entire finite element model. The hyperparameter were tuned to optimal values through the Bayesian algorithm, and the k-NN classification method showed the highest prediction rate and AUC-ROC among the tuned models. The 3D implementation method revealed higher performance than the 2D implementation method. The loading type data predicted through the machine learning model were mapped to the finite element model and comparatively verified through FE analysis. It was found that 3D implementation PIC bumper beam was superior to 2D implementation and uni-stacking sequence composite bumper.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.3
• /
• pp.1204-1227
• /
• 2020

The tradeoff between energy conservation and traffic balancing is a dilemma problem in Wireless Sensor Networks (WSNs). By analyzing the intrinsic relationship between cluster properties and long distance transmission energy consumption, we characterize three node sets of the cluster as a theoretical foundation to enhance high performance of WSNs, and propose optimal solutions by introducing rendezvous and Mobile Elements (MEs) to optimize energy consumption for prolonging the lifetime of WSNs. First, we exploit an approximate method based on the transmission distance from the different node to an ME to select suboptimal Rendezvous Point (RP) on the trajectory for ME to collect data. Then, we define data transmission routing sequence and model rendezvous planning for the cluster. In order to achieve optimization of energy consumption, we specifically apply the economic theory called Diminishing Marginal Utility Rule (DMUR) and create the utility function with regard to energy to develop an adaptive energy consumption optimization framework to achieve energy efficiency for data collection. At last, Rendezvous Transmission Algorithm (RTA) is proposed to better tradeoff between energy conservation and traffic balancing. Furthermore, via collaborations among multiple MEs, we design Two-Orbit Back-Propagation Algorithm (TOBPA) which concurrently handles load imbalance phenomenon to improve the efficiency of data collection. The simulation results show that our solutions can improve energy efficiency of the whole network and reduce the energy consumption of sensor nodes, which in turn prolong the lifetime of WSNs.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.22 no.3
• /
• pp.209-222
• /
• 1997

In this paper, we study a bargaining order problem where one buyer sequentially bargains with two sellers whose reservation prices are unknown to the buyer but correlated. Our main question is who the buyer should bargain first with to maximize his expected payoff. This type of problem is widely applicable to business and political situations where one party negotiates with multiple parties sequentially. One of the most important element in a sequential bargaining is "linkage effect" which exists when the aggreement of the previous bargaining affects the outcome of the following bargaining. To examine "linkage effect", we assume that the sellers'objects are similar so that the sellers' reservation prices are correlated. In addition, to consider incomplete information aspect regarding reservation prices, it is assumed that the sellers' reservation prices are unknown to the buyer. That is, we deal with one sided incomplete information case. In our model, there are two stages in each of which the buyer meets one seller. Since we are concerned with the bargaining order, we consider two different bargaining orders. Using game theory, we find a perfect Bayesian equilibrium and compute the buyer's expected payoff for each bargaining order. Finally we identify the advantageous bargaining order for the buyer by comparing the expected payoffs obtained under two different bargaining orders. Our results are as follows: the advantageous bargaining order depends on the prior probability of the seller type. However, in general, the buyer should bargain first with the seller whose object is less valuable to the buyer. The basic reason for our result is that the buyer wants to experiment in the first stage to find out the sellers' reservation prices and in doing so, to minimize the experimental cost and maximize potential gain in case of negotiation failure in the first stage. in the first stage.