• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.1
• /
• pp.1-10
• /
• 2022

Deep neural networks are widely used to solve various problems. In a fully connected neural network, the nonlinear activation function is a function that nonlinearly transforms the input value and outputs it. The nonlinear activation function plays an important role in solving the nonlinear problem, and various nonlinear activation functions have been studied. In this study, we propose a combined parametric activation function that can improve the performance of a fully connected neural network. Combined parametric activation functions can be created by simply adding parametric activation functions. The parametric activation function is a function that can be optimized in the direction of minimizing the loss function by applying a parameter that converts the scale and location of the activation function according to the input data. By combining the parametric activation functions, more diverse nonlinear intervals can be created, and the parameters of the parametric activation functions can be optimized in the direction of minimizing the loss function. The performance of the combined parametric activation function was tested through the MNIST classification problem and the Fashion MNIST classification problem, and as a result, it was confirmed that it has better performance than the existing nonlinear activation function and parametric activation function.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.9
• /
• pp.637-648
• /
• 2009

As the cost of disks decreases, PCs are soon expected to be equipped with a disk of 1TB or more. Assuming that a single person generates 1GB of data per month, 1TB is enough to store data for the entire lifetime of a person. This has lead to the growth of researches on lifelog management, which manages what people see and listen to in everyday life. Although many different lifelog management systems have been proposed, including those based on the relational data model, based on ontology, and based on file systems, they have all advantages and disadvantages: Those based on the relational data model provide good query processing performance but they do not support complex queries properly; Those based on ontology handle more complex queries but their performances are not satisfactory: Those based on file systems support only keyword queries. Moreover, these systems are lack of support for lifelog group management and do not provide a convenient user interface for modifying and adding tags (metadata) to lifelogs for effective lifelog search. To address these problems, we propose a lifelog management system based on the relational data model. The proposed system models lifelogs by using the relational data model and transforms queries on lifelogs into SQL statements, which results in good query processing performance. It also supports a simplified relationship query that finds a lifelog based on other lifelogs directly related to it, to overcome the disadvantage of not supporting complex queries properly. In addition, the proposed system supports for the management of lifelog groups by providing ways to create, edit, search, play, and share them. Finally, it is equipped with a tagging tool that helps the user to modify and add tags conveniently through the ion of various tags. This paper describes the design and implementation of the proposed system and its various applications.

• Asia pacific journal of information systems
• /
• v.24 no.2
• /
• pp.233-253
• /
• 2014

Can the stock market really be predicted? Stock market prediction has attracted much attention from many fields including business, economics, statistics, and mathematics. Early research on stock market prediction was based on random walk theory (RWT) and the efficient market hypothesis (EMH). According to the EMH, stock market are largely driven by new information rather than present and past prices. Since it is unpredictable, stock market will follow a random walk. Even though these theories, Schumaker [2010] asserted that people keep trying to predict the stock market by using artificial intelligence, statistical estimates, and mathematical models. Mathematical approaches include Percolation Methods, Log-Periodic Oscillations and Wavelet Transforms to model future prices. Examples of artificial intelligence approaches that deals with optimization and machine learning are Genetic Algorithms, Support Vector Machines (SVM) and Neural Networks. Statistical approaches typically predicts the future by using past stock market data. Recently, financial engineers have started to predict the stock prices movement pattern by using the SNS data. SNS is the place where peoples opinions and ideas are freely flow and affect others' beliefs on certain things. Through word-of-mouth in SNS, people share product usage experiences, subjective feelings, and commonly accompanying sentiment or mood with others. An increasing number of empirical analyses of sentiment and mood are based on textual collections of public user generated data on the web. The Opinion mining is one domain of the data mining fields extracting public opinions exposed in SNS by utilizing data mining. There have been many studies on the issues of opinion mining from Web sources such as product reviews, forum posts and blogs. In relation to this literatures, we are trying to understand the effects of SNS exposures of firms on stock prices in Korea. Similarly to Bollen et al. [2011], we empirically analyze the impact of SNS exposures on stock return rates. We use Social Metrics by Daum Soft, an SNS big data analysis company in Korea. Social Metrics provides trends and public opinions in Twitter and blogs by using natural language process and analysis tools. It collects the sentences circulated in the Twitter in real time, and breaks down these sentences into the word units and then extracts keywords. In this study, we classify firms' exposures in SNS into two groups: positive and negative. To test the correlation and causation relationship between SNS exposures and stock price returns, we first collect 252 firms' stock prices and KRX100 index in the Korea Stock Exchange (KRX) from May 25, 2012 to September 1, 2012. We also gather the public attitudes (positive, negative) about these firms from Social Metrics over the same period of time. We conduct regression analysis between stock prices and the number of SNS exposures. Having checked the correlation between the two variables, we perform Granger causality test to see the causation direction between the two variables. The research result is that the number of total SNS exposures is positively related with stock market returns. The number of positive mentions of has also positive relationship with stock market returns. Contrarily, the number of negative mentions has negative relationship with stock market returns, but this relationship is statistically not significant. This means that the impact of positive mentions is statistically bigger than the impact of negative mentions. We also investigate whether the impacts are moderated by industry type and firm's size. We find that the SNS exposures impacts are bigger for IT firms than for non-IT firms, and bigger for small sized firms than for large sized firms. The results of Granger causality test shows change of stock price return is caused by SNS exposures, while the causation of the other way round is not significant. Therefore the correlation relationship between SNS exposures and stock prices has uni-direction causality. The more a firm is exposed in SNS, the more is the stock price likely to increase, while stock price changes may not cause more SNS mentions.

• Geophysics and Geophysical Exploration
• /
• v.23 no.4
• /
• pp.243-252
• /
• 2020

To share an understanding of trajectory measurement in surveys using borehole, this tutorial summarizes the relevant mathematical principles of the borehole deviation survey based on coordinate transform. For uncased or open holes, calculations of the azimuth-deviation-tool face rotation using three-component accelerometer and magnetometer measurements are summarized. For the steel-cased holes, calculations are based on the time-derivative formula of the coordinate transform matrix; yaw-pitch-roll angles through time are mathematically determined by integrating the threecomponent angular velocity measurements from the gyroscope while also removing the Earth's rotation effect. Sensor and data fusion to increase the accuracy of borehole deviation survey is explained with an example of the method. These principles of borehole deviation surveys can be adapted for attitude estimation in air-borne surveys or for positioning in tunnels where global positioning system (GPS) signals cannot be accessed. Information on the optimization filter that must be incorporated in sensor fusion is introduced to help future research.

• Journal of The Korean Association of Information Education
• /
• v.20 no.5
• /
• pp.475-486
• /
• 2016

As software transforms the structure of industry, it becomes a key measure in determining market competitiveness. Therefore, various educational efforts have been attempted in Korea to cultivate software professionals to secure software competitiveness. While previous studies had focused mainly on the cognitive effectiveness of software education, the authors tried to focus on affective perspectives. The authors, therefore, aimed to analyze the predictive power of the recognition of software importance and learning flow on affective outcomes, such as efficacy of computational thinking skills, and attitude toward, and satisfaction with, software education. The data were collected from 103 sixth grade students who participated in a software education. Results show that software importance and learning flow had significant predictive power on affective outcomes; Learning flow mediated the relationship between software importance and affective outcomes. This study provides practical implications for improving affective outcomes in the design and implementation of software education.

• Journal of Internet Computing and Services
• /
• v.2 no.2
• /
• pp.71-93
• /
• 2001

Commonly to design a database system. a conceptual database has to be designed and then it is transformed into a logical database schema prior to building a target database system. This paper proposes a method which transforms a Temporal Entity-Relationship Model(TERM) into a Temporal Object-Oriented Model(TOOM) to build an efficient database schema. I formalize the time concept in view of object versioning and specify the constraints required during transformation procedure. The proposed transformation method contributes to getting the logical temporal data from the conceptual temporal events Without any loss of semantics, Compared to other approaches of supporting various properties, this approach is more general and efficient because it is the semantically seamless transformation method by using the orthogonality of types of objects, semantics of relationships and constraints over roles.

• Journal of KIISE:Databases
• /
• v.34 no.5
• /
• pp.420-436
• /
• 2007

The choice of an effective indexing method is crucial to guarantee the performance of the spatial join operator which is heavily used in geographical information systems. The $R^*$-tree based method is renowned as one of the most representative indexing methods. In this paper, we propose an efficient spatial join technique based on the DOT(Double Transformation) index, and compare it with the spatial Join technique based on the $R^*$-tree index. The DOT index transforms the MBR of an spatial object into a single numeric value using a space filling curve, and builds the $B^+$-tree from a set of numeric values transformed as such. The DOT index is possible to be employed as a primary index for spatial objects. The proposed spatial join technique exploits the regularities in the moving patterns of space filling curves to divide a query region into a set of maximal sub-regions within which space filling curves traverse without interruption. Such division reduces the number of spatial transformations required to perform the spatial join and thus improves the performance of join processing. The experiments with the data sets of various distributions and sizes revealed that the proposed join technique is up to three times faster than the spatial join method based on the $R^*$-tree index.

• The Journal of Korean Institute of Next Generation Computing
• /
• v.15 no.5
• /
• pp.64-74
• /
• 2019

Although a non-rigid registration has high demands in clinical practice, it has a high computational complexity and it is very difficult for ensuring the accuracy and robustness of registration. This study proposes a method of applying a non-rigid registration to 3D magnetic resonance images of brain in an unsupervised learning environment by using a deep-learning network. A feature vector between two images is produced through the network by receiving both images from two different patients as inputs and it transforms the target image to match the source image by creating a displacement vector field. The network is designed based on a U-Net shape so that feature vectors that consider all global and local differences between two images can be constructed when performing the registration. As a regularization term is added to a loss function, a transformation result similar to that of a real brain movement can be obtained after the application of trilinear interpolation. This method enables a non-rigid registration with a single-pass deformation by only receiving two arbitrary images as inputs through an unsupervised learning. Therefore, it can perform faster than other non-learning-based registration methods that require iterative optimization processes. Our experiment was performed with 3D magnetic resonance images of 50 human brains, and the measurement result of the dice similarity coefficient confirmed an approximately 16% similarity improvement by using our method after the registration. It also showed a similar performance compared with the non-learning-based method, with about 10,000 times speed increase. The proposed method can be used for non-rigid registration of various kinds of medical image data.

• The Journal of the Korea Contents Association
• /
• v.22 no.1
• /
• pp.391-404
• /
• 2022

Actor-network theory (ANT) pays attention to the relational effect between human and non-human actors, and transforms numerous networks between these actors by treating non-humans as human-like actors. This paper investigated various non-human actors related to the context before and after the introduction of NCS-based curriculum through ANT. This approach is because as a new system is introduced, the impact on the existing network and conflict situations can be looked at more closely. To this end, the researcher reviewed data from October 2014, when S College discussed whether to introduce an NCS-based curriculum, to February 2017, when practical operation was carried out and graduates were produced. In order to understand ANT theory, we analyzed based on the four stages of translation as claimed by Callon in the ANT theory. As a result, some meanings were confirmed in the case of reforming the curriculum of S College where the NCS-based curriculum was introduced. First, it is an in-depth analysis of the situation surrounding the curriculum, which has been overlooked by research on the existing curriculum. Second, it contributed to interpreting the 'hidden meaning' beyond the 'superficial meaning' of the curriculum within the university. Third, it was possible to indirectly check the conflicts and conflicts with the existing system that appeared in the process of introducing the new system to the College.

• Korean Journal of Culture and Social Issue
• /
• v.16 no.1
• /
• pp.43-61
• /
• 2010

In Korea, Pbichim refers to a psychological state caused by emotional damages that can occur within close relationships. In this state, one might feel reluctant to express one's feelings directly to the other party. It is also possible that Pbichim transforms into anger. This study is aimed to define the term Pbichim as an indigenous psychological concept. In Korea, it is common to express one's feelings indirectly and read the other party's inward thoughts. Pbichim reflects those cultural aspects. In order to examine the representation of Pbichim in Korea, we developed a questionnaire consisting of 15 open-ended questions. The participants were 119 undergraduate and graduate students at Korea University, and the data was analyzed qualitatively. As a result, four different aspects of Pbichim (unsatisfied expectation, being ignored, being alienated, and power struggle) could be differentiated by the situation in which people are likely to present Pbichim. The personality traits of Pbichim, the way of relieving it, as well as positive and negative functions of Pbichim were also elicited. In addition, it was found that Pbichim (the concept that has been negatively perceived) has an important function in maintaining and improving an interpersonal relationship in Korea. Lastly, the importance of mind reading within a certain cultural context is discussed.