• Toxicological Research
• /
• v.32 no.4
• /
• pp.289-300
• /
• 2016

Non-genotoxic carcinogens are substances that induce tumorigenesis by non-mutagenic mechanisms and long term rodent bioassays are required to identify them. Recent studies have shown that transcription profiling can be applied to develop early identifiers for long term phenotypes. In this study, we used rat liver expression profiles from the NTP (National Toxicology Program, Research Triangle Park, USA) DrugMatrix Database to construct a gene classifier that can distinguish between non-genotoxic carcinogens and other chemicals. The model was based on short term exposure assays (3 days) and the training was limited to oxidative stressors, peroxisome proliferators and hormone modulators. Validation of the predictor was performed on independent toxicogenomic data (TG-GATEs, Toxicogenomics Project-Genomics Assisted Toxicity Evaluation System, Osaka, Japan). To build our model we performed Random Forests together with a recursive elimination algorithm (VarSelRF). Gene set enrichment analysis was employed for functional interpretation. A total of 770 microarrays comprising 96 different compounds were analyzed and a predictor of 54 genes was built. Prediction accuracy was 0.85 in the training set, 0.87 in the test set and increased with increasing concentration in the validation set: 0.6 at low dose, 0.7 at medium doses and 0.81 at high doses. Pathway analysis revealed gene prominence of cellular respiration, energy production and lipoprotein metabolism. The biggest target of toxicogenomics is accurately predict the toxicity of unknown drugs. In this analysis, we presented a classifier that can predict non-genotoxic carcinogenicity by using short term exposure assays. In this approach, dose level is critical when evaluating chemicals at early time points.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.31 no.1 s.256
• /
• pp.26-35
• /
• 2007

The vehicle of urban transit is a complex system that consists of various electric, electronic, and mechanical equipments, and the maintenance cost of this complex and large-scale system generally occupies sixty percent of the LCC (Life Cycle Cost). For reasonable establishing of maintenance strategies, safety security and cost limitation must be considered at the same time. The concept of system reliability has been introduced and optimized as the key of reasonable maintenance strategies. For optimization, three preceding studies were accomplished; standardizing a maintenance classification, constructing RBD (Reliability Block Diagram) of VVVF (Variable Voltage Variable Frequency) urban transit, and developing a web based reliability evaluation system. Historical maintenance data in terms of reliability index can be derived from the web based reliability evaluation system. In this paper, we propose applying inverse problem analysis method and hybrid neuro-genetic algorithm to system reliability optimization for using historical maintenance data in database of web based system. Feed-forward multi-layer neural networks trained by back propagation are used to find out the relationship between several component reliability (input) and system reliability (output) of structural system. The inverse problem can be formulated by using neural network. One of the neural network training algorithms, the back propagation algorithm, can attain stable and quick convergence during training process. Genetic algorithm is used to find the minimum square error.

• Journal of Korean Society of Rural Planning
• /
• v.17 no.1
• /
• pp.89-98
• /
• 2011

Recently, the food industry has evolved into a new and innovative trend according to its globalization and change of food consumption patterns. However, it is hard for the traditional food industry in Korea to meet the changing consumers' needs because of its poorer quality control and lower industrialization technology than other advanced industries. Also the knowledges acquired through a lot of time and efforts would be lost after the human resources with tacit knowledges leave by their too much aging. Especially, the 21st century would be called as knowledge based society which means that knowledge be the important contributing factor in the economic growth. In this regard, this study aimed at proposing the knowledge based system for systematically managing or preserving knowledges of Mosisongpyun industry in Yeonggwang County to seek for the sustainable development of the traditional food industry in Korea. The knowledge based system of Mosisongpyun industry in Yeonggwang County is finally proposed as follows; First, hardware is composed with the necessary unit facilities such as interpretive center, learning and experience room, library, etc. And the integrating facilities such as Mosisongpyun theme park, traditional village, and knowledge industrialization support center are proposed. Second, software is composed with the necessary unit softwares such as the preservation manual of traditional knowledge and skill, web-site administrator, development of graded textbooks, development database software, etc. And the integrating softwares such as development of innovation and management ability in Mosisongpyun industry are proposed. Third, humanware is composed with the necessary unit programs such as exhibition, own training program, incubator support system, etc. And the integrating programs such as the farm association corporation, the testing and research institute, the institution of learning and training are proposed.

• Journal of The Korean Astronomical Society
• /
• v.52 no.6
• /
• pp.217-225
• /
• 2019

We apply a modified Convolutional Neural Network (CNN) model in conjunction with transfer learning to predict whether an active region (AR) would produce a ≥C-class or ≥M-class flare within the next 24 hours. We collect line-of-sight magnetogram samples of ARs provided by the SHARP from May 2010 to September 2018, which is a new data product from the HMI onboard the SDO. Based on these AR samples, we adopt the approach of shuffle-and-split cross-validation (CV) to build a database that includes 10 separate data sets. Each of the 10 data sets is segregated by NOAA AR number into a training and a testing data set. After training, validating, and testing our model, we compare the results with previous studies using predictive performance metrics, with a focus on the true skill statistic (TSS). The main results from this study are summarized as follows. First, to the best of our knowledge, this is the first time that the CNN model with transfer learning is used in solar physics to make binary class predictions for both ≥C-class and ≥M-class flares, without manually engineered features extracted from the observational data. Second, our model achieves relatively high scores of TSS = 0.640±0.075 and TSS = 0.526±0.052 for ≥M-class prediction and ≥C-class prediction, respectively, which is comparable to that of previous models. Third, our model also obtains quite good scores in five other metrics for both ≥C-class and ≥M-class flare prediction. Our results demonstrate that our modified CNN model with transfer learning is an effective method for flare forecasting with reasonable prediction performance.

• Journal of the Korean Data and Information Science Society
• /
• v.25 no.2
• /
• pp.373-384
• /
• 2014

Data mining is to clarify pattern or correlation of mass data of complicated structure and to predict the diverse outcomes. This technique is used in the fields of finance, telecommunication, circulation, medicine and so on. In this paper, we selected risk factors of respiratory diseases in the field of medicine. The data we used was divided into respiratory diseases group and health group from the Gyeongsangbuk-do database of Community Health Survey conducted in 2012. In order to select major risk factors, we applied data mining techniques such as neural network, logistic regression, Bayesian network, C5.0 and CART. We divided total data into training and testing data, and applied model which was designed by training data to testing data. By the comparison of prediction accuracy, CART was identified as best model. Depression, smoking and stress were proved as the major risk factors of respiratory disease.

• Journal of Practical Engineering Education
• /
• v.12 no.1
• /
• pp.23-29
• /
• 2020

Data generated by the development of ICT, the diversification of ICT devices and services and the expansion of social media are categorized as big data characterized by the amount, variety and speed of the data. The spread of the use of big data is expected to have the effects of identifying the status quo by analyzing data in all industries, predicting the future, and creating opportunities to apply it. However, while it is imperative for these things to be done, the nation still lacks professional training institutions or curricula. In this case study, we will investigate and compare the state of education for the training of big data personnel in Korea, find out what level and level of education is being trained to nurture balanced professionals, and prepare an opportunity to think about how it can help students create value at a time when the need for education is growing in the wake of awareness of big data.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.13 no.3
• /
• pp.1-13
• /
• 2010

The aim of this study is to analyze gold-silver mineral potential in the Taebaeksan mineralized district, Korea using a Geographic Information System(GIS) and an artificial neural network(ANN) model. A spatial database considering Au and Ag deposit, geology, fault structure and geochemical data of As, Cu, Mo, Ni, Pb and Zn was constructed for the study area using the GIS. The 46 Au and Ag mineral deposits were randomly divided into a training set to analyze mineral potential using ANN and a test set to verify mineral potential map. In the ANN model, training sets for areas with mineral deposits and without them were selected randomly from the lower 10% areas of the mineral potential index derived from existing mineral deposits using likelihood ratio. To support the reliability of the Au-Ag mineral potential map, some of rock samples were selected in the upper 5% areas of the mineral potential index without known deposits and analyzed for Au, Ag, As, Cu, Pb and Zn. As the result, No. 4 of sample exhibited more enrichments of all elements than the others.

• Economic and Environmental Geology
• /
• v.35 no.1
• /
• pp.67-74
• /
• 2002

The purpose of this study is to determine the weights of the factors for landslide susceptibility analysis using artificial neural network. Landslide locations were identified from interpretation of aerial photographs, field survey data, and topography. The landslide-related factors such as topographic slope, topographic curvature, soil drainage, soil effective thickness, soil texture, wood age and wood diameter were extracted from the spatial database in study area, Yongin. Using these factors, the weights of neural networks were calculated by backpropagation training algorithm and were used to determine the weight of landslide factors. Therefore, by interpreting the weights after training, the weight of each landslide factor can be ranked based on its contribution to the classification. The highest weight is topographic slope that is 5.33 and topographic curvature and soil texture are 1 and 1.17, respectively. Weight determination using backprogpagation algorithms can be used for overlay analysis of GIS so the factor that have low weight can be excluded in future analysis to save computation time.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.6
• /
• pp.773-778
• /
• 2009

This paper studied the training methods less affected by the emotional variation for the development of the robust speech recognition system. For this purpose, the effect of emotional variations on the speech signal were studied using speech database containing various emotions. The performance of the speech recognition system trained by using the speech signal containing no emotion is deteriorated if the test speech signal contains the emotions because of the emotional difference between the test and training data. In this study, it is observed that vocal tract length of the speaker is affected by the emotional variation and this effect is one of the reasons that makes the performance of the speech recognition system worse. In this paper, vocal tract normalization method is used to develop the robust speech recognition system for emotional variations. Experimental results from the isolated word recognition using HMM showed that the vocal tract normalization method reduced the error rate of the conventional recognition system by 41.9% when emotional test data was used.

• Journal of the Korean Geosynthetics Society
• /
• v.19 no.1
• /
• pp.11-23
• /
• 2020

In this paper development artificial neural networks (ANN) for preliminary design and prediction of urban tunnelling and deep excavation-induced ground settlement was presented. In order to form training and validation data sets for the ANN development, field design and measured data were collected for various tunnelling and deep-excavation sites. The field data were then used as a database for the ANN training. The developed ANN was validated against a testing set and the unused field data in terms of statistical parameters such as R², RMSE, and MAE. The practical use of ANN was demonstrated by applying the developed ANN to hypothetical conditions. It was shown that the developed ANN can be effectively used as a tool for preliminary excavation design and ground settlement prediction for urban excavation problems.