• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.1-17
• /
• 2017

The deep learning framework is software designed to help develop deep learning models. Some of its important functions include "automatic differentiation" and "utilization of GPU". The list of popular deep learning framework includes Caffe (BVLC) and Theano (University of Montreal). And recently, Microsoft's deep learning framework, Microsoft Cognitive Toolkit, was released as open-source license, following Google's Tensorflow a year earlier. The early deep learning frameworks have been developed mainly for research at universities. Beginning with the inception of Tensorflow, however, it seems that companies such as Microsoft and Facebook have started to join the competition of framework development. Given the trend, Google and other companies are expected to continue investing in the deep learning framework to bring forward the initiative in the artificial intelligence business. From this point of view, we think it is a good time to compare some of deep learning frameworks. So we compare three deep learning frameworks which can be used as a Python library. Those are Google's Tensorflow, Microsoft's CNTK, and Theano which is sort of a predecessor of the preceding two. The most common and important function of deep learning frameworks is the ability to perform automatic differentiation. Basically all the mathematical expressions of deep learning models can be represented as computational graphs, which consist of nodes and edges. Partial derivatives on each edge of a computational graph can then be obtained. With the partial derivatives, we can let software compute differentiation of any node with respect to any variable by utilizing chain rule of Calculus. First of all, the convenience of coding is in the order of CNTK, Tensorflow, and Theano. The criterion is simply based on the lengths of the codes and the learning curve and the ease of coding are not the main concern. According to the criteria, Theano was the most difficult to implement with, and CNTK and Tensorflow were somewhat easier. With Tensorflow, we need to define weight variables and biases explicitly. The reason that CNTK and Tensorflow are easier to implement with is that those frameworks provide us with more abstraction than Theano. We, however, need to mention that low-level coding is not always bad. It gives us flexibility of coding. With the low-level coding such as in Theano, we can implement and test any new deep learning models or any new search methods that we can think of. The assessment of the execution speed of each framework is that there is not meaningful difference. According to the experiment, execution speeds of Theano and Tensorflow are very similar, although the experiment was limited to a CNN model. In the case of CNTK, the experimental environment was not maintained as the same. The code written in CNTK has to be run in PC environment without GPU where codes execute as much as 50 times slower than with GPU. But we concluded that the difference of execution speed was within the range of variation caused by the different hardware setup. In this study, we compared three types of deep learning framework: Theano, Tensorflow, and CNTK. According to Wikipedia, there are 12 available deep learning frameworks. And 15 different attributes differentiate each framework. Some of the important attributes would include interface language (Python, C ++, Java, etc.) and the availability of libraries on various deep learning models such as CNN, RNN, DBN, and etc. And if a user implements a large scale deep learning model, it will also be important to support multiple GPU or multiple servers. Also, if you are learning the deep learning model, it would also be important if there are enough examples and references.

• Korean Journal of Soil Science and Fertilizer
• /
• v.38 no.3
• /
• pp.164-171
• /
• 2005

The aim of this study was to evaluate the load of nutrient from paddy fields. Water management practices that can reduce eutrophication and meet water quality requirements will also be addressed. Continuous monitoring from May to September in 2002 and 2003 was conducted for water quantification and qualification at the intensive paddy fields in Icheon, Gyunggi province of Korea. Water balance and concentration variation of nitrogen and phosphorus in the water were independently compared for water quality assessment at each rice cultivation period. Rice land preparation and transplanting periods usually marked the highest water demand when compared to other periods of cultivation. Overall, a greater net irrigation ratio was observed during the transplanting period in 2002 (92.3%) and 2003 (87.2%). The measured total N loads of precipitation, irrigation, drainage, and percolation during the rice cultivation period were 9.9, 41.6, 22.1, and $5.5kg\;ha^{-1}$ for 2002 and 15.8, 55.4, 17.3, and $7.5kg\;ha^{-1}$ for 2003, respectively. The measured total P loads of precipitation, irrigation, drainage, and percolation during the rice cultivation period were 2.1, 13.0, 3.6, and $1.8kg\;ha^{-1}$ for 2002 and 1.6, 15.0, 5.0, and $1.2kg\;ha^{-1}$ for 2003, respectively. Daily nutrient load followed the pattern of surface drainage water, but this pattern was changed by rainfall events. The nutrient load in drainage water depends on rainfall and surface drainage water amount from the paddy fields. Interestingly, the load of total N and total P output was smaller than the input load due to the natural infiltration that Occurred during the rice cultivation period. It is concluded that the paddy fields have a beneficial effect on the ecosystem and can reduce eutrophication in the water.

• Journal of Korea Technology Innovation Society
• /
• v.13 no.1
• /
• pp.68-98
• /
• 2010

This study was conducted to analyze the results of the implementation of next-generation growth engine program that was promoted across all government agencies for five years from 2004 as part of a range of initiatives aimed at expanding the nation's economic growth potential. The results were analyzed in this study using five indices: program purpose and design; strategic planning; program management; program results; and pan-governmental agency cooperation and coordination. The results of the study include the following. First, regarding program purpose and design, it was found that consistent leadership in the promotion of the programs was insufficient as the relevant program promotion systems and financial resources were dispersed among governmental agencies, even though the objectives and validity of the programs were recognized. Second, with regard to strategic planning, it was found that although the program objectives and technical development strategy had been established at the beginning of the program, they were biased toward the technical objectives and mainly implemented by the technology suppliers. Third, regarding program management, it was found that the responsibility for general administration, ranging from task planning to policy improvement, was given to the appointed program director but that the system of cooperation among the agencies was insufficient to carry out the relevant tasks. Fourth, regarding the results of the program, it was difficult to understand the results consistently as the economic objectives were not clearly presented, even though the technical objectives were achieved despite the short implementation period of the program. Fifth, with regard to pan-governmental agency cooperation and coordination, it was found that the coordination organization whose remit was to implement the program was established pursuant to the Basic Law on Science and Technology, but that no detailed regulations or guidelines on the operation of the organization were drawn up. To efficiently plan and execute future pan-governmental agency R&D programs that are similar to the next-generation new growth engine program, various requirements should be met, namely, 1) joint planning and consistent program design among governmental agencies, 2) clarification of the program objectives and budget allocation system, 3) establishment of a pan-governmental agency program operation and assessment system, 4) formulation of a strategy for linking R&D with standardization, and 5) enactment of pan-governmental agency joint operation rules.

• Clinical and Experimental Pediatrics
• /
• v.48 no.8
• /
• pp.846-856
• /
• 2005

Purpose : To measure the peak myocardial tissue velocities and patterns of longitudinal motion of atrioventricular(AV) annuli and assess body weight and heart rates-related changes in normal children. Methods : Using pulsed wave Tissue Doppler Imaging(TDI), we measured peak systolic, early and late diastolic myocardial velocities in 72 normal children at six different sites in apical-4 chamber (A4C) view and at four different sites in apical-2 chamber(A2C) view and compared those values with each other, also observing effects with body weights and heart rates. Longitudinal motions of the AV annuli were measured at three different sites in A4C. Results : There were no significant differences of the TDI parameters between gender, ECHO-machines and among the three Doctors performing TDI. Peak myocardial velocities were significantly higher at the base of the heart than in the mid-ventricular region and in the right lateral ventricular wall than in the left lateral ventricular wall or IVS. The TDI parameters showed no significant correlation with fractional shortening(%). Peak systolic and early diastolic myocardial velocities had no correlation with heart rates, but peak late diastolic velocities and A/E ratio correlated positively with heart rates. Correlations between the TDI parameters and body weight were inconsistent. Absolute longitudinal displacement and % displacement were not differ between gender and not correlated with the TDI parameters. Conclusion : We measured the peak myocardial velocities with TDI and the longitudinal motion of the AV annuli using M-mode echocardiography in normal children. With more large scale evaluation, we may establish reference values in normal children and broaden clinical applicabilities in congenital and acquired heart diseases.

• Journal of the Korean Regional Science Association
• /
• v.35 no.3
• /
• pp.45-58
• /
• 2019

The government is recently pushing for five large-scale public housing sites near Seoul, the so-called third wave of new towns. In this regard, this study sought to analyze the key considerations in planning new towns from the perspective of the residents who are the consumers, using survey data. For this purpose, frequency analysis and variance analysis(ANOVA) were conducted on existing first and second era of new towns and the third era of new towns scheduled for construction. Eight indicators, such as environmental comfort, self-sufficiency such as jobs, transportation access and convenience, were set as subordinate variables, and characteristics of residents, including gender, age, number of households, household income, occupation, and residential areas, were set as explanatory variables. According to the analysis, the respondents rated the first era of new towns more positively than the second era of new towns. For self-sufficiency items such as jobs, both the first and second era of new towns showed low levels. In addition, for the eight indicators, the first era of new towns were no significant differences depending on gender, number of households, or household income, and the second era of new towns were no significant differences by the number of households, household income, occupations or place of residence. However, for new towns in the third period, the assessment of the importance of each indicator by age, number of households, household income and occupation showed significant differences. The results of the multi-comparison analysis of the third era of new towns showed that the importance of environmental comfort was highly valued by the youth, the managerial/professional/clerical position, single or five more persons of household, and the youth, high income household, the managerial/professional/clerical position when it comes to accessibility and convenience of transportation. It suggests that various personal characteristics and demands for each of the planning indicators need to be considered in planning for the third era of new towns.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.40 no.5
• /
• pp.19-32
• /
• 2012

The purpose of this study is to quantitatively measure the view target and the view area through the selection of landscape control point by selecting Jimyeong-dong Dong-gu and Yeongyeong-dong Buk-gu Daegu, which are planned as the large scale housing complex development area, as the target places. It is very meaningful that from the simulation based on this measurement, the improvement methods are attempted to be suggested at the project level. The results of this study are as follows. First, the total number of viewing targets derived from the literature analysis for the selection of the viewing targets was 24, and finally derived main viewing targets were 4 places. Second, the total number of selection criteria of the derived landscape control point was 15, and these were re-categorized as prospect, accessibility and publicness according to the common property. The preliminary LCPs were selected by measuring the publicness and accessibility, and because of the said measurement, a total of 43 preliminary LCPs were selected. The final LCPs were selected by estimating the prospect of the selected preliminary LCPs, and as a result of estimation, a total of 29 final LCPs were selected. Finally, the total number of evaluation indicators derived from literature analysis was 26. Because of the valuation by the landscape control point, it was found that the 2 view areas were the I grade, 3 areas were II grade and 3 areas were the V grade, the lowest grade among 29 view areas. From the analysis on problems for the improvement methods, 4 improvement-indicators including the diversity of land mosaic were selected for the view area-1 without considering the development project drawing. In addition, for the view area-2 with considering the development project drawing, the landscape as the scenery forests was well formed, and the arrangement of architectures for the security of view corridor was right angle arrangement, and their floor number was 10.

• Journal of Korean Society of societal Security
• /
• v.4 no.2
• /
• pp.73-78
• /
• 2011

In order to resolve the problem of change in global climate which is worsening as days go by and to preemptively cope with strengthened restriction on carbon emission, the government enacted 'Framework Act on Low Carbon Green Growth' in 2010 and selected green technology and green industry as new national growth engines. For this reason, the necessity to use the un-utilized waste heat across the whole industrial system has become an issue, and studies on and applications of recycling in the agricultural and fishery fields such as cultivation of tropical crops and flatfishes by utilizing the waste heat and thermal effluent generated by large industrial complexes including power plants are being actively carried out. In this study, we looked into the domestic and overseas examples of having utilized waste heat abandoned in the form of power plant thermal effluent, and carried out economic efficiency evaluation of sturgeon aquaculture utilizing thermal effluent of Yeongwol LNG Combined Cycle Power Plant in Gangwon-do. In this analysis, we analyzed the economic efficiency of a model business plan divided into three steps, starting from a small scale in order to minimize the investment risk and financial burden, which is then gradually expanded. The business operation period was assumed to be 10 years (2012~2021), and the NVP (Net Present Value) and economic efficiency (B/C) for the operation period (10 years) were estimated for different loan size by dividing the size of external loan by stage into 80% and 40% based on the basic statistics secured through a site survey. Through the result of analysis, we can see that reducing the size of the external loan is an important factor in securing greater economic efficiency as, while the B/C is 1.79 in the case the external loan is 80% of the total investment, it is presumed to be improved to 1.81 when the loan is 40%. As the findings of this study showed that the economic efficiency of sturgeon aquaculture utilizing thermal effluent of power plant can be secured, it is presumed that regional development project items with high added value can be derived though this, and, in addition, this study will greatly contribute to reinforcement of the capability of local governments to cope with climate change.

• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.75-95
• /
• 2020

With the growth of social networks, various forms of SNS have emerged. Based on various motivations for use such as interactivity, information exchange, and entertainment, SNS users are also on the fast-growing trend. Facebook is the main SNS channel, and companies have started using Facebook pages as a public relations channel. To this end, in the early stages of operation, companies began to secure a number of fans, and as a result, the number of corporate Facebook fans has recently increased to as many as millions. from a corporate perspective, Facebook is attracting attention because it makes it easier for you to meet the customers you want. Facebook provides an efficient advertising platform based on the numerous data it has. Advertising targeting can be conducted using their demographic characteristics, behavior, or contact information. It is optimized for advertisements that can expose information to a desired target, so that results can be obtained more effectively. it rethink and communicate corporate brand image to customers through contents. The study was conducted through Facebook advertising data, and could be of great help to business people working in the online advertising industry. For this reason, the independent variables used in the research were selected based on the characteristics of the content that the actual business is concerned with. Recently, the company's Facebook page operation goal is to go beyond securing the number of fan pages, branding to promote its brand, and further aiming to communicate with major customers. the main figures for this assessment are Facebook's 'OK', 'Attachment', 'Share', and 'Number of Click' which are the dependent variables of this study. in order to measure the outcome of the target, the consumer's response is set as a key measurable key performance indicator (KPI), and a strategy is set and executed to achieve this. Here, KPI uses Facebook's ad numbers 'reach', 'exposure', 'like', 'share', 'comment', 'clicks', and 'CPC' depending on the situation. in order to achieve the corresponding figures, the consideration of content production must be prior, and in this study, the independent variables were organized by dividing into three considerations for content production into three. The effects of content material, content structure, and message styles on Facebook's user behavior were analyzed using regression analysis. Content materials are related to the content's difficulty, company relevance, and daily involvement. According to existing research, it was very important how the content would attract users' interest. Content could be divided into informative content and interesting content. Informational content is content related to the brand, and information exchange with users is important. Interesting content is defined as posts that are not related to brands related to interesting movies or anecdotes. Based on this, this study started with the assumption that the difficulty, company relevance, and daily involvement have an effect on the dependent variable. In addition, previous studies have found that content types affect Facebook user activity. I think it depends on the combination of photos and text used in the content. Based on this study, the actual photos were used and the hashtag and independent variables were also examined. Finally, we focused on the advertising message. In the previous studies, the effect of advertising messages on users was different depending on whether they were narrative or non-narrative, and furthermore, the influence on message intimacy was different. In this study, we conducted research on the behavior that Facebook users' behavior would be different depending on the language and formality. For dependent variables, 'OK' and 'Full Click Count' are set by every user's action on the content. In this study, we defined each independent variable in the existing study literature and analyzed the effect on the dependent variable, and found that 'good' factors such as 'self association', 'actual use', and 'hidden' are important. Could. Material difficulties', 'actual participation' and 'large scale * difficulties'. In addition, variables such as 'Self Connect', 'Actual Engagement' and 'Sexual Sexual Attention' have been shown to have a significant impact on 'Full Click'. It is expected that through research results, it is possible to contribute to the operation and production strategy of company Facebook operators and content creators by presenting a content strategy optimized for the purpose of the content. In this study, we defined each independent variable in the existing research literature and analyzed its effect on the dependent variable, and we could see that factors on 'good' were significant such as 'self-association', 'reality use', 'concernal material difficulty', 'real-life involvement' and 'massive*difficulty'. In addition, variables such as 'self-connection', 'real-life involvement' and 'formative*attention' were shown to have significant effects for 'full-click'. Through the research results, it is expected that by presenting an optimized content strategy for content purposes, it can contribute to the operation and production strategy of corporate Facebook operators and content producers.