• Journal of Korea Multimedia Society
• /
• v.23 no.5
• /
• pp.625-632
• /
• 2020

In the diagnosis of lung cancer, the tumor size is measured by the longest diameter of the tumor in the entire slice of the CT. In order to accurately estimate the size of the tumor, it is better to measure the volume, but there are some limitations in calculating the volume in the clinic. In this study, we propose an algorithm to segment lung cancer by applying a custom loss function that combines focal loss and dice loss to a U-Net model that shows high performance in segmentation problems in chest CT images. The combination of values of the various parameters in custom loss function was compared to the results of the model learned. The purposed loss function showed F1 score of 88.77%, precision of 87.31%, recall of 90.30% and average precision of 0.827 at α=0.25, γ=4, β=0.7. The performance of the proposed custom loss function showed good performance in lung cancer segmentation.

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2003.10a
• /
• pp.95-100
• /
• 2003

In this paper, we propose solutions to resolve the problem of many spelling variants and the problem of lack of annotated corpus for training, which are two among the main difficulties in named entity recognition in biomedical domain. To resolve the problem of spotting valiants, we propose a use of edit-distance as a feature for SVM. And we propose a use of virtual examples to automatically expand the annotated corpus to resolve the lack-of-corpus problem. Using virtual examples, the annotated corpus can be extended in a fast, efficient and easy way. The experimental results show that the introduction of edit-distance produces some improvements in protein name recognition performance. And the model, which is trained with the corpus expanded by virtual examples, outperforms the model trained with the original corpus. According to the proposed methods, we finally achieve the performance 75.80 in F-measure(71.89% in precision,80.15% in recall) in the experiment of protein name recognition on GENIA corpus (ver.3.0).

• Journal of the Korean Society for Precision Engineering
• /
• v.4 no.2
• /
• pp.47-55
• /
• 1987

Recently, in precision working, precision is in submicron. Therefore, when we measure various finished goods in superfine measurement, because it is relatively difficult to disregard effect of surroundings, these effect of surroundings must be compensated or canceled. In this study, for roundness measurement, three point method is researched which is able to cancel the effect of rotation accuracy of axis and eccenricity of workpiece. It is difference between this three point method and tradi- tional three point method whose measuring apparatus have three movable pickups posit- ioned with angle and between the pickups. As a results, when rotation accuracy of axis is varied from $0.02\mu\textrm{m}$ to $0.05\mu\textrm{m}$ the width of variation of measured roundness is $0.04\mu\textrm{m}$. And, when eccentricity of workpiece is varied from 0 to $4\mu\textrm{m}$, the width of variation of measured roundness is $0.005\mu\textrm{m}$. These error width are disregardable because they are in 10% of measured roundness. Therefore, by this three point method, the effect of rotation accuracy of axis and the effect of eccentricity of workpiece are canceled. And we are able to select the angle between the pickups ($\phi$ and $\tau$) by means of relation between $F_{k}$ and K.

• Toxicological Research
• /
• v.17
• /
• pp.41-45
• /
• 2001

The methods used for risk assessment from exposure to chemicals are well established. in most cases where toxicity other than carcinogenesis is being considered, the standard method relies on establishing the No Observed Adverse Effect Level (NOAEL) in the most sensitive animal toxicity study and using an appropriate safety factor (SF) to determine the exposure which would be associated with an acceptable risk. For carcinogens a different approach is used because it has been argued there is no threshold of effect. Thus mathematical equations are used to extrapolate from the high doses used in ani-mal experiments. These methods have been strongly criticised in recent years on several grounds. The most cogent criticisms are a) the equations are not based on a thorough understanding of the mechanisms of carcinogenesis and b) the outcome of a risk assessment based on such models varies more as a consequence of changes to the assumptions and equation used than it does from the data derived from carcinogenicity experiments. Other criticisms include the absence of any measure of the variance on the risk assessment and the selection of default values that are very conservative. Recent advances in the application of risk assessment emphasise that measures of both the exposure and the hazard should be considered as a distribution of values. The outcome of such a risk assessment provides an estimate of the distribution of the risks.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.48 no.3
• /
• pp.63-71
• /
• 2006

A downdraft gasifier was made of stainless steel for biomass gasification. Internal reactor had a 300 mm diameter and 8 air intakes. Three thermocouples were installed to measure the temperature inside the reactor. Three different biomass fuels were provided in the experiments to find out the effects of fuel conditions on gasification processes; charcoals, woodchips, and mixture of woodchip and charcoals. Two different experiments were conducted fer charcoal experiments, small and larger sizes of charcoal fuels. It took about 10 minutes after ignition to generate combustible producer gas when charcoal was f9d, but 20 or more minutes for woodchips. When the gasification was stabilized, the highest temperature was observed just below the combustion zone. The air flow rate for woodchip experiment was provided at 25% of a stoichiometric requirement of combustion, which was within the range of typical air flow rate fer woody biomass gasification. Carbon monoxide concentrations were also within the values reported in the previous studies, ranging 20 to 30% depending on fuel types. It could be seen that fuel size and heating value were very important parameters in biomass gasification. These parameters should be taken into account in operating and designing biomass gasifiers.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.60 no.2
• /
• pp.76-81
• /
• 2011

Power system frequency is the main index of power quality indicating an abnormal state and disturbances of systems. The nominal frequency is deviated by sudden change in generation and load or faults. Power system is used as frequency relay to detection for off-nominal frequency operation and connecting a generator to an electrical system, and V/F relay to detection for an over-excitation condition. Under these circumstances, power system should maintain the nominal frequency. And frequency and frequency deviation should accurately measure and quickly estimate by frequency measurement device. The well-known classical method, frequency estimation technique based on the DFT, could be produce the gain error in accuracy. To meet the requirements for high accuracy, recently Wavelet transforms and analysis are receiving new attention. The Wavelet analysis is possible to calculate the time-frequency analysis which is easy to obtain frequency information of signals. However, it is difficult to apply in real-time implementation because of heavy computation burdens. Nowadays, the computational methods using the Wavelet function and transformation techniques have been searched on these fields. In this paper, we apply the Recursive Discrete Wavelet Transform (RDWT) for the frequency estimation. In order to evaluate performance of the proposed technique, the user-defined arbitrary waveforms are used.

• Journal of Korean Institute of Industrial Engineers
• /
• v.43 no.3
• /
• pp.192-202
• /
• 2017

This study develops a text spam filtering system for Facebook based on two variable categories: keywords learned from Instagram and meta-information of Facebook posts. Since there is no explicit labels for spam/ham posts, we utilize hash tags in Instagram to train classification models. In addition, the filtering accuracy is enhanced by considering meta-information of Facebook posts. To verify the proposed filtering system, we conduct an empirical experiment based on a total of 1,795,067 and 761,861 Facebook and Instagram documents, respectively. Employing random forest as a base classification algorithm, experimental result shows that the proposed filtering system yield 99% and 98% in terms of filtering accuracy and F1-measure, respectively. We expect that the proposed filtering scheme can be applied other web services suffering from massive spam posts but no explicit spam labels are available.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.11
• /
• pp.5594-5615
• /
• 2019

Malicious social robots, which are disseminators of malicious information on social networks, seriously affect information security and network environments. The detection of malicious social robots is a hot topic and a significant concern for researchers. A method based on classification has been widely used for social robot detection. However, this method of classification is limited by an unbalanced data set in which legitimate, negative samples outnumber malicious robots (positive samples), which leads to unsatisfactory detection results. This paper proposes the use of generative adversarial networks (GANs) to extend the unbalanced data sets before training classifiers to improve the detection of social robots. Five popular oversampling algorithms were compared in the experiments, and the effects of imbalance degree and the expansion ratio of the original data on oversampling were studied. The experimental results showed that the proposed method achieved better detection performance compared with other algorithms in terms of the F1 measure. The GAN method also performed well when the imbalance degree was smaller than 15%.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.7
• /
• pp.2468-2487
• /
• 2015

In web searching, trustable and precise results are greatly affected by the inherent uncertainty in the input queries. Queries submitted to search engines are by nature ambiguous and constitute a significant proportion of the instances given to web search engines. Ambiguous queries pose real challenges for the web search engines due to versatility of information. Temporal based approaches whereas somehow reduce the uncertainty in queries but still lack to provide results according to users aspirations. Web search science has created an interest for the researchers to incorporate contextual information for resolving the uncertainty in search results. In this paper, we propose an Adaptive Disambiguation Approach (ADA) of hybrid nature that makes use of both the temporal and contextual information to improve user experience. The proposed hybrid approach presents the search results to the users based on their location and temporal information. A Java based prototype of the systems is developed and evaluated using standard dataset to determine its efficacy in terms of precision, accuracy, recall, and F1-measure. Supported by experimental results, ADA demonstrates better results along all the axes as compared to temporal based approaches.

• Korean Journal of Human Ecology
• /
• v.11 no.2
• /
• pp.189-196
• /
• 2002

The purpose of this study were to elicit the attitudes toward underwear, to compare the difference according to the demographic variables, and to find the relationship between the attitudes toward underwear and social-psychological characteristics of university female students in Cheongju city. The questionnaires were to measure the attitudes toward underwear, the demographic information, and social-psychological characteristics. As statistical analysis, frequencies, mean, t-test, F-test, factor analysis, and correlation were used. The results were as follows: 1) The female students regarded tactile sensation of underwear fabrics as the important element in their attitude toward underwear, while they showed a little interest about gorgeous underwear style. 2) The freshman students have strongly. fixed ideas about the under wear, the other(sophomore, junior, senior) students showed more interest in decorative and fashionable aspects of underwear. 3) The attitudes toward underwear were classifies into six factors; intension about high quality, decorative design, stereotype, immodesty, style and comfort. 4) There were partially significant differences according to demographic variables in the attitudes toward underwear. 5) Also, there were significant correlations between the social-psychological characteristics and the attitudes toward underwear.