• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.2
• /
• pp.78.1-78.1
• /
• 2019

There are a huge number of faint objects that have not been observed due to the lack of large and deep surveys. In this study, we demonstrate that a deep learning approach can produce a better quality deep image from a single pass imaging so that could be an alternative of conventional image stacking technique or the expensive large and deep surveys. Using data from the Sloan Digital Sky Survey (SDSS) stripe 82 which provide repeatedly scanned imaging data, a training data set is constructed: g-, r-, and i-band images of single pass data as an input and r-band co-added image as a target. Out of 151 SDSS fields that have been repeatedly scanned 34 times, 120 fields were used for training and 31 fields for validation. The size of a frame selected for the training is 1k by 1k pixel scale. To avoid possible problems caused by the small number of training sets, frames are randomly selected within that field each iteration of training. Every 5000 iterations of training, the performance were evaluated with RMSE, peak signal-to-noise ratio which is given on logarithmic scale, structural symmetry index (SSIM) and difference in SSIM. We continued the training until a GAN model with the best performance is found. We apply the best GAN-model to NGC0941 located in SDSS stripe 82. By comparing the radial surface brightness and photometry error of images, we found the possibility that this technique could generate a deep image with statistics close to the stacked image from a single-pass image.

• Journal of Intelligence and Information Systems
• /
• v.19 no.2
• /
• pp.39-54
• /
• 2013

Recent explosive increase of electronic commerce provides many advantageous purchase opportunities to customers. In this situation, customers who do not have enough knowledge about their purchases, may accept product recommendations. Product recommender systems automatically reflect user's preference and provide recommendation list to the users. Thus, product recommender system in online shopping store has been known as one of the most popular tools for one-to-one marketing. However, recommender systems which do not properly reflect user's preference cause user's disappointment and waste of time. In this study, we propose a novel recommender system which uses data mining and multi-model ensemble techniques to enhance the recommendation performance through reflecting the precise user's preference. The research data is collected from the real-world online shopping store, which deals products from famous art galleries and museums in Korea. The data initially contain 5759 transaction data, but finally remain 3167 transaction data after deletion of null data. In this study, we transform the categorical variables into dummy variables and exclude outlier data. The proposed model consists of two steps. The first step predicts customers who have high likelihood to purchase products in the online shopping store. In this step, we first use logistic regression, decision trees, and artificial neural networks to predict customers who have high likelihood to purchase products in each product group. We perform above data mining techniques using SAS E-Miner software. In this study, we partition datasets into two sets as modeling and validation sets for the logistic regression and decision trees. We also partition datasets into three sets as training, test, and validation sets for the artificial neural network model. The validation dataset is equal for the all experiments. Then we composite the results of each predictor using the multi-model ensemble techniques such as bagging and bumping. Bagging is the abbreviation of "Bootstrap Aggregation" and it composite outputs from several machine learning techniques for raising the performance and stability of prediction or classification. This technique is special form of the averaging method. Bumping is the abbreviation of "Bootstrap Umbrella of Model Parameter," and it only considers the model which has the lowest error value. The results show that bumping outperforms bagging and the other predictors except for "Poster" product group. For the "Poster" product group, artificial neural network model performs better than the other models. In the second step, we use the market basket analysis to extract association rules for co-purchased products. We can extract thirty one association rules according to values of Lift, Support, and Confidence measure. We set the minimum transaction frequency to support associations as 5%, maximum number of items in an association as 4, and minimum confidence for rule generation as 10%. This study also excludes the extracted association rules below 1 of lift value. We finally get fifteen association rules by excluding duplicate rules. Among the fifteen association rules, eleven rules contain association between products in "Office Supplies" product group, one rules include the association between "Office Supplies" and "Fashion" product groups, and other three rules contain association between "Office Supplies" and "Home Decoration" product groups. Finally, the proposed product recommender systems provides list of recommendations to the proper customers. We test the usability of the proposed system by using prototype and real-world transaction and profile data. For this end, we construct the prototype system by using the ASP, Java Script and Microsoft Access. In addition, we survey about user satisfaction for the recommended product list from the proposed system and the randomly selected product lists. The participants for the survey are 173 persons who use MSN Messenger, Daum Caf$\acute{e}$, and P2P services. We evaluate the user satisfaction using five-scale Likert measure. This study also performs "Paired Sample T-test" for the results of the survey. The results show that the proposed model outperforms the random selection model with 1% statistical significance level. It means that the users satisfied the recommended product list significantly. The results also show that the proposed system may be useful in real-world online shopping store.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09b
• /
• pp.49-52
• /
• 2003

In this research, we propose a mechanism to develop an inference engine and expert systems based on relational database and SQL (structured query language). Generally, former researchers had tried to develop an expert systems based on text-oriented knowledge base and backward/forward (chaining) inference engine. In these researches, however, the speed of inference was remained as a tackling point in the development of agile expert systems. Especially, the forward inference needs more times than backward inference. In addition, the size of knowledge base, complicate knowledge expression method, expansibility of knowledge base, and hierarchies among rules are the critical limitations to develop an expert systems. To overcome the limitations in speed of inference and expansibility of knowledge base, we proposed a relational database-oriented knowledge base and forward inference engine. Therefore, our proposed mechanism could manipulate the huge size of knowledge base efficiently, and inference with the large scaled knowledge base in a short time. To this purpose, we designed and developed an SQL-based forward inference engine using relational database. In the implementation process, we also developed a prototype expert system and presented a real-world validation data set collected from medical diagnosis field.

• 한국전산유체공학회:학술대회논문집
• /
• 2001.05a
• /
• pp.189-194
• /
• 2001

In this paper, non-reacting and reacting flowfields were computed using a preconditioned Navier-Stokes solver. The preconditioning technique of Merkle et al. and TVD scheme or Chakravarthy and Osher was employed and the results obtained using developed code have a good agreement with the previous results and experimental data. The preconditioned Wavier-Stokes equation set with low Reynolds number $\kappa-\epsilon$ equation and species continuity equations, are discretized with strongly implicit manner and time integrated with LU-SSOR scheme. For the purpose of treating unsteady problem the duel-time stepping scheme was employed. For the validation of the code in incompressible flow regime, steady driven square cavity flow was considered and calculation result shows reasonably good agreement with the result of incompressible code. Shock wave/boundary layer interaction problem was considered to show the shock capturing performance of preconditioned-TVD scheme. To validate unsteady flow, acoustic oscillation problem was calculated, and supersonic premix flame of $H_2$-air reaction problem which is calculated with turbulence model, 9-species/18-reaction step reaction model, shows reasonable agreement with the previous results. As a result, the preconditioning method has an advantage to calculate incompressible and compressible flow through one code and preconditioned solver easily developed from standard compressible code with minor efforts. But additional computational time and computer memory is required due to preconditioning matrix.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.15 no.5
• /
• pp.69-74
• /
• 2015

In this paper the patterns of information system is reduced so that control rules can guarantee fast response of queries in database. Generally an information system includes many kinds of necessary and unnecessary attribute. In particular, inconsistent information system is less likely to acquire the accuracy of response. Hence we are interested in the simple and understandable rules that can represent useful patterns by means of rough entropy and Bayesian posterior probability. We propose an algorithm which can reduce control rules to a minimum without inadequate patterns such that the implication between condition attributes and decision attributes is measured through the framework of rough entropy. Subsequently the validation of the proposed algorithm is showed through test information system of new employees appointment.

• Proceedings of the KSR Conference
• /
• 2010.06a
• /
• pp.1798-1805
• /
• 2010

The Railroad Safety Technology R&D Program has been supported by the Ministry of Land, Transport and Maritime Affairs. The Program consists of dozen or more projects and thus is quite complex, which requires effective project management. This paper is discussing how to effectively manage the railroad safety technology R&D program based on systems engineering management plan (SEMP). The key issues dealt in this paper include the SE management, requirements management, verification and validation, integrated data base management, and traceability management. To achieve the goal, we first defined an appropriate SE process to be adopted in the program and the result was documented in the systems engineering management plan. According to the process defined, we set up the environmental frame in database. Using the database and SEMP, each project of the program was verified with respect to the corresponding requirements utilizing the traceability. All these efforts were carried out using a computer-aided SE tool, which enables efficient management of complex database.

• 한국전산유체공학회:학술대회논문집
• /
• 2008.03b
• /
• pp.268-271
• /
• 2008

If a reactor vessel is failed to retain a molten corium in a postulated severe accident, the molten corium is released outside the reactor vessel into a reactor cavity. The molten corium would attack the concrete wall and basemat of the reactor cavity, which may lead to inevitable concrete decompositions and possible radiological releases. In the OECD/MCCI project, a series of tests were performed to secure the data for cooling the molten corium spread out at the reactor cavity and for the long-term CCI (Core Concrete Interaction). Also, a MCCI (Molten Core Concrete Interaction) analysis code, CORQUENCH was upgraded at Argonne National Laboratory with embedding the new models developed for the tests. This paper deals with analyses of MCCI at plant scale under the conditions of top flooding using the upgraded CORQUENCH code. The modeling approach is briefly summarized first, followed by presentation of a validation calculation that illustrates the predicative capability of the modeling tool. With this background in place, the model is then used to carry out a parametric set of scoping calculations that define approximate coolability envelopes for the LCS (Limestone Common Sand) concrete that has been evaluated in the OECD/MCCI project.

• Proceedings of the Korea Technical Association of the Pulp and Paper Industry Conference
• /
• 2006.06b
• /
• pp.365-372
• /
• 2006

In this paper, the possibility of statistical modeling from the pulp and peroxide bleaching condition variables to predict optical properties of a bleached chemimechanical pulp used in a newsprint paper machine at Mazandaran Wood and Paper Industries Company (MWPI) was studied. Due to the variations in the opacity and the brightness of the bleached pulp at MWPI and to tackle this problem, it was decided to study the possibility of modeling the bleaching process. To achieve this purpose, Multi-variate Regression Analysis was used for model building and it was found that there is a relationship between independent variables and pulp brightness as well as pulp opacity, consequently, two models were constructed. Then, model validation was carried out using new data set in the bleaching plant at MWPI to test model predictive ability and its performance.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.10
• /
• pp.4080-4097
• /
• 2020

In recent years, convolutional neural network (CNN) has become the primary method for face detection. But its shortcomings are obvious, such as expensive calculation, heavy model, etc. This makes CNN difficult to use on the mobile devices which have limited computing and storage capabilities. Therefore, the design of lightweight CNN for face detection is becoming more and more important with the popularity of smartphones and mobile Internet. Based on the CPU real-time face detector FaceBoxes, we propose a multi-task lightweight face detector, which has low computing cost and higher detection precision. First, to improve the detection capability, the squeeze and excitation modules are used to extract attention between channels. Then, the textual and semantic information are extracted by shallow networks and deep networks respectively to get rich features. Finally, the landmark detection module is used to improve the detection performance for small faces and provide landmark data for face alignment. Experiments on AFW, FDDB, PASCAL, and WIDER FACE datasets show that our algorithm has achieved significant improvement in the mean average precision. Especially, on the WIDER FACE hard validation set, our algorithm outperforms the mean average precision of FaceBoxes by 7.2%. For VGA-resolution images, the running speed of our algorithm can reach 23FPS on a CPU device.

• Ocean Systems Engineering
• /
• v.1 no.1
• /
• pp.31-57
• /
• 2011

Impact pressure due to sloshing is of great concern for the ship owners, designers and builders of the LNG carriers regarding the safety of LNG containment system and hull structure. Sloshing of LNG in partially filled tank has been an active area of research with numerous experimental and numerical investigations over the past decade. In order to accurately predict the sloshing impact load, a new numerical method was developed for accurate resolution of violent sloshing flow inside a three-dimensional LNG tank including wave breaking, jet formation, gas entrapping and liquid-gas interaction. The sloshing flow inside a membrane-type LNG tank is simulated numerically using the Finite-Analytic Navier-Stokes (FANS) method. The governing equations for two-phase air and water flows are formulated in curvilinear coordinate system and discretized using the finite-analytic method on a non-staggered grid. Simulations were performed for LNG tank in transverse and longitudinal motions including horizontal, vertical, and rotational motions. The predicted impact pressures were compared with the corresponding experimental data. The validation results clearly illustrate the capability of the present two-phase FANS method for accurate prediction of impact pressure in sloshing LNG tank including violent free surface motion, three-dimensional instability and air trapping effects.