• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.395-397
• /
• 2019

Despite technical advance, human error is the main reason for maritime accidents. To ensure a safety of maritime transporting environment, technical and methodological improvement to react to various types of maritime accidents should be developed instead of ambiguously anticipating maritime accidents due to human errors. Survey, questionnaires, and interview have been routinely applied to understand objective human lookout pattern differences in various navigational situations. Although the descriptive methodology helps systematically categorizing different patterns of human behavior to avoid accidents, the subjective methods limit to objectively recognize physical behavior patterns during navigation. The purpose of the study is to develop an objective lookout pattern detection system using wearable sensors in the simulated navigation environment. In the simulated maritime navigation environment, each participant performed a given navigational situation by wearing the wearable sensors on the wrist, trunk, and head. Activity classification algorithm that was developed in the previous navigation activity classification research was applied. The physical lookout behavior patterns before and after situation-aware showed distinctive patterns, and the results are expected to reduce human errors of navigators.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.9 no.1
• /
• pp.75-82
• /
• 1991

The objective of this paper is to investigate the optimal Robust estimation and scale estimator that could be used to treat the gross errors in a self calibrating bundle adjustment. In order to test the variability in performance of the different weighting schemes in accurately detecting gross error, five robust estimation methods and three types of scale estimators were used. And also, two difference control point patterns(high density control, sparse density control) and three types of gross errors(4$\sigma o$, 20$\sigma o$, 50$\sigma o$) were used for comparison analysis. As a result, Anscombe's robust estimation produced the best results in accuracy among the robust estimation methods considered. when considering the scale estimator about control point patterns, It can be seen that Type II scale estimator provided the best accuracy in high density control pattern. On the other hand, In the case of sparse density control pattern, Type III scale estimator showed the best results in accuracy. Therefore it is expected to apply to robustified bundle adjustment using the optimal scale estimator which can be used for eliminating the gross error in precise structure analysis.

• KIPS Transactions on Computer and Communication Systems
• /
• v.9 no.1
• /
• pp.9-16
• /
• 2020

The k-anonymity scheme has been widely used to protect private information when Big Data are distributed to a third party for research purposes. When the scheme is applied, an optimal k value determination is one of difficult problems to be resolved because many factors should be considered. Currently, the determination has been done almost manually by human experts with their intuition. This leads to degrade performance of the anonymization, and it takes much time and cost for them to do a task. To overcome this problem, a simple idea has been proposed that is based on machine learning. This paper describes implementations and experiments to realize the proposed idea. In thi work, a deep neural network (DNN) is implemented using tensorflow libraries, and it is trained and tested using input dataset. The experiment results show that a trend of training errors follows a typical pattern in DNN, but for validation errors, our model represents a different pattern from one shown in typical training process. The advantage of the proposed approach is that it can reduce time and cost for experts to determine k value because it can be done semi-automatically.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.13 no.8
• /
• pp.3660-3665
• /
• 2012

For blind equalization, it is necessary to open an eye pattern quickly in the early stage of equalization, after that it is important to lower an error level of equalizer output signal. This paper discusses coarse error estimation using signal points specifically determined and fine error estimation using original signal constellation, and proposes two suggestions for how to take advantage of the two error estimation methods. The two error estimates, respectively, are effective to quickly open an eye pattern in the state of eye pattern closed, or to lower the level of an error in the steady-state after the eye pattern opening. Two blind equalization algorithms are proposed and their performances are compared, which select one of the two error estimates depending on the state of convergence of the equalizer, or combine two errors weightedly according to the relative reliabilities of the two error estimates, and calculate the new error.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.31 no.11C
• /
• pp.1030-1036
• /
• 2006

In this paper, we present two methods of correcting bit errors in constant amplitude multi-code (CAMC) CDMA, which uses the redundant bits only. The first method is a parity-based bit correction with hard-decision, where the received signals despread into n two-dimensional structure with both horizontal parity and vertical parity. Then, an erroneous bit is corrected for each $4{\times}4$ pattern. The second method is a turbo decoding, which is modified from the decoding of a single parity check product code (SPCPC). Experimental results show that, in the second method, the redundant bits in CAMC can be fully used for the error correction and so they are not really a loss of channel bandwidth. Hence, CAMC provides both a low peak-to-average power ratio and robustness to bit errors.

• Journal of Korean Institute of Industrial Engineers
• /
• v.26 no.3
• /
• pp.193-199
• /
• 2000

In many studies, considerable attention has been focussed upon choosing a model which represents underlying process of time series and forecasting the future. In the real world, however, there may be some cases that one model can not reflect all the characteristics of original time series. Under such circumstances, we may get better performance by combining the forecasts from several models. The most popular methods for combining forecasts involve taking a weighted average of multiple forecasts. But the weights are usually unstable. In cases the assumptions of normality and unbiasedness for forecast errors are satisfied, a Bayesian method can be used for updating the weights. In the real world, however, there are many circumstances the Bayesian method is not appropriate. This paper proposes a PNN(Probabilistic Neural Net) approach as a method for combining forecasts that can be applied when the assumption of normality or unbiasedness for forecast errors is not satisfied. In this paper, PNN method, which is similar to Bayesian approach, is suggested as an updating method of the unstable weights in the combination of the forecasts. The PNN method has been usually used in the field of pattern recognition. Unlike the Bayesian approach, it requires no assumption of a specific prior distribution because it gets probabilities by using the distribution estimated from given data. Empirical results reveal that the PNN method offers superior predictive capabilities.

• Korean Journal of Cognitive Science
• /
• v.15 no.3
• /
• pp.15-24
• /
• 2004

The perceived position of a flashing target object is generally biased towards the direction of eye movement when there is no reference around the target. Current research examined the localization accuracy of a flashing target relative to a static reference. The perceived location of the target relative to the reference was distorted and the pattern of perceptual distortion systematically depended on the position of the reference relative to the target. This kind of result was consistently observed regardless of the distance between the reference and the target and direction of pursuit eye movement. We have discussed how these results could he explained by the theories previously suggested to explain the localization of objects.

• Journal of Nutrition and Health
• /
• v.19 no.5
• /
• pp.333-341
• /
• 1986

Several aspects including physical development, reflex acquistion, neuromotor development and learning behavior at Y water maze were compared at the progeny of rats fed low 91.2mg/kg diet) or adequate leves(22mg/kg diet 0 of pyridoxine during growth, gestation, lactation, and adult period. Physical development and development of reflexes (righting reflex, cliff avoidance, negative geotaxis, palmar grasp, and startle reflex to sound) appeared different between control and deficient groups but not significantly. At the 2nd week, rats spent more time in supported standing during 6 minute period was longer in the control then the deficient groups. In the Y-water maze position reversal test, learning ability as judged by the number of errors was not different among three groups, but the rats in supplemented group(DC) reached the escape platform in significantly shorter time than the other two groups, which may suggest their emotional instability. In the visual discrimination test, the performance of rats from the supplemented group had the lower errors than the other groups on the early test days. but as the testing period progressed, the performance of rats in the supplemented group became inferior to those of the control and deficient groups. The performance of control group became superior to that of the deficient group.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.9
• /
• pp.159-165
• /
• 1994

In this paper, we propose a multi-layer associative neural network structure suitable for hardware implementaion with the function of performance refinement and improved robutst capability. Unlike other methods which reduce network complexity by putting restrictions on synaptic weithts, we are imposing a requirement of hidden layer neurons for the function. The proposed network has synaptic weights obtainted by Hebbian rule between adjacent layer's memory patterns such as Kosko's BAM. This network can be extended to arbitary multi-layer network trainable with Genetic algorithm for getting hidden layer memory patterns starting with initial random binary patterns. Learning is done to minimize newly defined network error. The newly defined error is composed of the errors at input, hidden, and output layers. After learning, we have bidirectional recall process for performance improvement of the network with one-shot recall. Experimental results carried out on pattern recognition problems demonstrate its performace according to the parameter which represets relative significance of the hidden layer error over the sum of input and output layer errors, show that the proposed model has much better performance than that of Kosko's bidirectional associative memory (BAM), and show the performance increment due to the bidirectionality in recall process.

• Phonetics and Speech Sciences
• /
• v.7 no.2
• /
• pp.29-38
• /
• 2015

This study explores reading fluency among elementary school students considering language level and family income(low SES). Forty eight students from 1st to 3rd grades participated in two paragraph reading tasks. Half of the children were from low income family and half of the children had low lexical knowledge. Reading fluency as in the number of correctly read syllables per minute, the total error frequency and error types were used to compare group differences. There were significant differences in the number of correctly read syllables per minute between two income groups and two language groups. There was a significant difference between low income group and non-low income group in total number of errors only when children's lexical knowledge were low. There were no group differences in error types of repetition and omission. Substitution and insertion error seemed to reflect the total error pattern. These results imply the importance of early screening and early involvement for children with low lexical knowledge from low income family. Monitoring and early intervention will support these children's reading development.