• Journal of the Korean GEO-environmental Society
• /
• v.9 no.6
• /
• pp.61-70
• /
• 2008

In this study, undrained triaxial (CU) tests were performed on low-plastic silt of Nakdong River in order to investigate the undrained shear behavior of low-plastic silt. In experimental results, the deviator stress showed the hardening behavior after reaching its yield stress like the tendency of common sand, and the pore water pressure was gradually decreased to critical state after the maximum value. In the effective stress paths, regardless of consolidation stress or overconsolidation ratios, both a critical state line (CSL) and a phase transformation line (PTL) exist in the effective stress path that is similar to the case of sand. The behavior of low-plastic silt was predicted by the Modified Cam-Clay (MCC) model, the Jordan and the Elman-jordan model that is artificial neural network model. According to predicted results, the overall undrained shear behavior of low-plastic silt could not be predicted with the MCC model, but the Jordan and Elman-Jordan model showed well-matched experiment results.

• Tunnel and Underground Space
• /
• v.14 no.6 s.53
• /
• pp.399-410
• /
• 2004

The utilization of AS(Artificial Intelligence) and Database could be considered as an useful access for the application of underground information from the point of a geotechnical methodology. Its detailed usage has been recently studied in many fields of geo-sciences. In this paper, the target of usage is on controlling the injection of grout which more scientific access is needed in the grouting that has been used a major method in many engineering application. As the proposals for this problem it is suggested the methodology consisting of a fuzzy-neural hybrid system and a database. The database was firstly constructed for parameters dynamically varied according to the conditions of rock mass during the injection of grout. And then the conceptional model for the fuzzy-neural hybrid system was investigated fer optimally finding the controlling range of the grout valve. The investigated model applied to four cases, and it is found that the controlling range of the grout valve was reasonably deduced corresponding to the mechanical phenomena occurred by the injection of grout. Consequently, the algorithm organizing the fuzzy-neural hybrid system and the database as a system can be considered as a tool for controlling the injection condition of grout.

• Journal of Korean Society of Transportation
• /
• v.23 no.7 s.85
• /
• pp.149-158
• /
• 2005

The object of this paper is to develop the shortest path algorithm. The existing shortest path algorithm models are developed while considering travel time and travel distance. A few problems occur in these shortest path algorithm models, which have paid no regard to cognition of users, such as when user who doesn't have complete information about the trip meets a strange road or when the route searched from the shortest path algorithm model is not commonly used by users in real network. This paper develops a shortest path algorithm model to provide ideal route that many people actually prefer. In order to provide the ideal shortest path with the consideration of travel time, travel distance and road cognition, travel time is predicted by using Kalman filtering and travel distance is predicted by using GIS attributions. The road cognition is considered by using space data of GIS. Optimal routes provided from this paper are shortest distance path, shortest time path, shortest path considering distance and cognition and shortest path considering time and cognition.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.31 no.10A
• /
• pp.960-967
• /
• 2006

Mobile communication is evolving toward next generation system that is capable to provide ubiquitous human-centric services with giving no restriction on time and location. In particular, context-awareness will be realized using various sensors and devices and related technologies under the ubiquitous Next Generation (NG) mobile communications environment where user's demand and surrounding context are changing dynamically. By such context awareness, context aware service which provides best suitable services for a user by analyzing user's needs and situational information, will be one of the promising NG mobile services. In this article, we propose NG mobile service environment with focusing on generic service traffic flows supported by the NG service platform which gives various means of connecting users and service providers. Then, we propose evolutional phases of context-aware services based on ranges of context information and evolutional trend of telecommunication technologies and networking environment in NG mobile service environment.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.19 no.5
• /
• pp.187-194
• /
• 2019

Traditional demand forecasting methods are difficult to meet the needs of companies due to rapid changes in the market and the diversification of individual consumer needs. In a diversified production environment, the right demand forecast is an important factor for smooth yield management. Many of the existing predictive models commonly used in industry today are limited in function by little. The proposed model is designed to overcome these limitations, taking into account the part where each model performs better individually. In this paper, variables are extracted through Gray Relational analysis suitable for dynamic process analysis, and statistically predicted data is generated that includes characteristics of historical demand data produced through ARIMA forecasts. In combination with the LSTM model, demand forecasts can then be calculated by reflecting the many factors that affect demand forecast through an architecture that is structured to avoid the long-term dependency problems that the neural network model has.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.4
• /
• pp.189-198
• /
• 2024

Lipreading is one of the important parts of speech recognition, and several studies have been conducted to improve the performance of lipreading in lipreading systems for speech recognition. Recent studies have used method to modify the model architecture of lipreading system to improve recognition performance. Unlike previous research that improve recognition performance by modifying model architecture, we aim to improve recognition performance without any change in model architecture. In order to improve the recognition performance without modifying the model architecture, we refer to the cues used in human lipreading and set other regions such as chin and cheeks as regions of interest along with the lip region, which is the existing region of interest of lipreading systems, and compare the recognition rate of each region of interest to propose the highest performing region of interest In addition, assuming that the difference in normalization results caused by the difference in interpolation method during the process of normalizing the size of the region of interest affects the recognition performance, we interpolate the same region of interest using nearest neighbor interpolation, bilinear interpolation, and bicubic interpolation, and compare the recognition rate of each interpolation method to propose the best performing interpolation method. Each region of interest was detected by training an object detection neural network, and dynamic time warping templates were generated by normalizing each region of interest, extracting and combining features, and mapping the dimensionality reduction of the combined features into a low-dimensional space. The recognition rate was evaluated by comparing the distance between the generated dynamic time warping templates and the data mapped to the low-dimensional space. In the comparison of regions of interest, the result of the region of interest containing only the lip region showed an average recognition rate of 97.36%, which is 3.44% higher than the average recognition rate of 93.92% in the previous study, and in the comparison of interpolation methods, the bilinear interpolation method performed 97.36%, which is 14.65% higher than the nearest neighbor interpolation method and 5.55% higher than the bicubic interpolation method. The code used in this study can be found a https://github.com/haraisi2/Lipreading-Systems.

• Journal of Intelligence and Information Systems
• /
• v.16 no.1
• /
• pp.93-116
• /
• 2010

Ensemble approach is applied to the detection modeling of illegal cash accommodation (ICA) that is the well-known type of fraudulent usages of credit cards in far east nations and has not been addressed in the academic literatures. The performance of fraud detection model (FDM) suffers from the imbalanced data problem, which can be remedied to some extent using an ensemble of many classifiers. It is generally accepted that ensembles of classifiers produce better accuracy than a single classifier provided there is diversity in the ensemble. Furthermore, recent researches reveal that it may be better to ensemble some selected classifiers instead of all of the classifiers at hand. For the effective detection of ICA, we adopt ensemble size reduction technique that prunes the ensemble of all classifiers using accuracy and diversity measures. The diversity in ensemble manifests itself as disagreement or ambiguity among members. Data imbalance intrinsic to FDM affects our approach for ICA detection in two ways. First, we suggest the training procedure with over-sampling methods to obtain diverse training data sets. Second, we use some variants of accuracy and diversity measures that focus on fraud class. We also dynamically calculate the diversity measure-Forward Addition and Backward Elimination. In our experiments, Neural Networks, Decision Trees and Logit Regressions are the base models as the ensemble members and the performance of homogeneous ensembles are compared with that of heterogeneous ensembles. The experimental results show that the reduced size ensemble is as accurate on average over the data-sets tested as the non-pruned version, which provides benefits in terms of its application efficiency and reduced complexity of the ensemble.

• Journal of the HCI Society of Korea
• /
• v.5 no.2
• /
• pp.7-14
• /
• 2010

Urban Data collected through Sensor Network is becoming crucial to understand and analyse a city. Thus, the Ubiquitous Sensor Network builds the foundation of the u-City development. This research aims to develop an energy monitoring application with an intuitive visualization environment which integrates energy usage information on top of urban geospatial information. Such a system will be able to facilitate effective energy supply plan at the early stages of urban planning, and eventually to encourage citizens to conserve energy by giving them real time monitoring information in an easy to understand visual environment. The system provides multiple layers of energy-related information coupled with the geospatial information layer in order to accommodate multiple viewpoints. On the other hand, the system provides logical Level of Detail control based on urban spatial information hierarchy. We defined the system concept and functions, and designed the data structure and the methods of information visualization. This paper presents the visualization methods, data structure, interactions scenarios which combines spacial information, E-GIS data and the energy related sensor data. Furthermore this research tries to introduce the concept of Social Sensor Networking to enhance the monitoring quality.