• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.8
• /
• pp.667-675
• /
• 2013

This paper presents vision-based techniques for underwater landmark detection, map-based localization, and SLAM (Simultaneous Localization and Mapping) in structured underwater environments. A variety of underwater tasks require an underwater robot to be able to successfully perform autonomous navigation, but the available sensors for accurate localization are limited. A vision sensor among the available sensors is very useful for performing short range tasks, in spite of harsh underwater conditions including low visibility, noise, and large areas of featureless topography. To overcome these problems and to a utilize vision sensor for underwater localization, we propose a novel vision-based object detection technique to be applied to MCL (Monte Carlo Localization) and EKF (Extended Kalman Filter)-based SLAM algorithms. In the image processing step, a weighted correlation coefficient-based template matching and color-based image segmentation method are proposed to improve the conventional approach. In the localization step, in order to apply the landmark detection results to MCL and EKF-SLAM, dead-reckoning information and landmark detection results are used for prediction and update phases, respectively. The performance of the proposed technique is evaluated by experiments with an underwater robot platform in an indoor water tank and the results are discussed.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.44 no.5
• /
• pp.41-53
• /
• 2002

This study was conducted to derive the regional design rainfall by the regional frequency analysis based on the regionalization of the precipitation suggested by the first report of this project. According to the regions and consecutive durations, optimal design rainfalls were derived by the regional frequency analysis for L-moment in the second report of this project. Using the LH-moment ratios and Kolmogorov-Smirnov test, the optimal regional probability distribution was identified to be the Generalized extreme value (GEV) distribution among applied distributions. regional and at-site parameters of the GEV distribution were estimated by the linear combination of the higher probability weighted moments, LH-moment. Design rainfall using LH-moments following the consecutive duration were derived by the regional and at-site analysis using the observed and simulated data resulted from Monte Carlo techniques. Relative root-mean-square error (RRMSE), relative bias (RBIAS) and relative reduction (RR) in RRMSE for the design rainfall were computed and compared in the regional and at-site frequency analysis. Consequently, it was shown that the regional analysis can substantially more reduce the RRMSE, RBIAS and RR in RRMSE than at-site analysis in the prediction of design rainfall. Relative efficiency (RE) for an optimal order of L-moments was also computed by the methods of L, L1, L2, L3 and L4-moments for GEV distribution. It was found that the method of L-moments is more effective than the others for getting optimal design rainfall according to the regions and consecutive durations in the regional frequency analysis. Diagrams for the design rainfall derived by the regional frequency analysis using L-moments were drawn according to the regions and consecutive durations by GIS techniques.

• Journal of Convergence for Information Technology
• /
• v.8 no.1
• /
• pp.161-171
• /
• 2018

This study set out to identify crime vulnerable areas with the GIS spatial analysis technique for the prediction of crimes. Crime vulnerable areas were extracted from the statistics of crimes with the GIS hotspot analysis technique and the inverse distance weighted(IDW) method applied to different crimes according to places and use districts. The scope of surveillance and weight were calculated for each of CPTED surveillance elements including CCTV, streetlamp, patrol division, and police substation. Maps of crime vulnerable areas were overlapped one after another to make a CPTED-based one expressed in four grades(safety, attention, warning, and risk).

• Imaging Science in Dentistry
• /
• v.44 no.4
• /
• pp.293-299
• /
• 2014

Purpose: This study was performed to find the relationship between pain and joint effusion using magnetic resonance imaging (MRI) in temporomandibular disorder (TMD) patients. Materials and Methods: The study subjects included 232 TMD patients. The inclusion criteria in this study were the presence of spontaneous pain or provoked pain on one or both temporomandibular joints (TMJs). The provoked pain was divided into three groups: pain on palpation (G1), pain on mouth opening (G2), and pain on mastication (G3). MRI examinations were performed using a 1.5-T MRI scanner. T1- and T2-weighted images with para-sagittal and para-coronal images were obtained. According to the T2-weighted image findings, the cases of effusions were divided into four groups: normal, mild (E1), moderate (E2), and marked effusion (E3). A statistical analysis was carried out using the $X^2$ test with SPSS (version 12.0, SPSS Inc., Chicago, IL, USA). Results: Spontaneous pain, provoked pain, and both spontaneous and provoked pain were significantly related to joint effusion in TMD patients (p<0.05). However, among the various types of provoked pain, pain on palpation of the masticatory muscles and TMJ (G1) was not related to joint effusion in TMD patients (p>0.05). Conclusion: Spontaneous pain was related to the MRI findings of joint effusion; however, among the various types of provoked pain, pain on palpation of the masticatory muscles and TMJ was not related to the MRI findings of joint effusion. These results suggest that joint effusion has a significant influence on the prediction of TMJ pain.

• Journal of Internet Computing and Services
• /
• v.10 no.4
• /
• pp.115-126
• /
• 2009

A software size estimation has to be analyzed in the beginning of the software life-cycle and helpful to the prediction of its size and cost. The software cost has been calculated by estimating software size from the user's point of view since the function point method based on international standards was introduced for the estimation of software size in 2004. However, the current function point method is not easy to be exploited for unfamiliar user, and has a problem that it cannot estimate the proper size for software such as engineering software, scientific calculations and simulation with complicated internal computational logic. This paper presents an improved model which can simplify the existing function point measurement procedure, and perform the estimation of software size in easy and fast way at the initial stage of project. Moreover, it presents a mathematical weighted value calculation model which can solve the problem of the fixed complexity weighted value and reflect the characteristics of organization as its data is pilled up. Our evaluation shows that the presented model has advantage that it can measure the size more rapidly than the existing FPA methods and has more correlation with LOC.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.9
• /
• pp.59-68
• /
• 2022

A recommender system covers users, searches the items or services which users will like, and let users purchase them. Because recommendations from a recommender system are predictions of users' preferences for the items which they do not purchase yet, it is rarely possible to be drawn a perfect answer. An evaluation has been conducted to determine whether a prediction is right or not. However, it can be lower user's satisfaction if a recommender system focuses on only the preferences, that is caused by a 'filter bubble effect'. The filter bubble effect is an algorithmic bias that skews or limits the information an individual user sees on the recommended list. It is the reason why multiple metrics are required to evaluate recommender systems, and a diversity metrics is mainly used for it. In this paper, we compare three different methods for enhancing diversity for personalized recommendation - bin packing, weighted random choice, greedy re-ranking - with a practical e-commerce data acquired from a fashion shopping mall. Besides, we present the difference between experimental results and F1 scores.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.8
• /
• pp.1-10
• /
• 2023

Recently, NAND flash memories have replaced magnetic disks due to non-volatility, high capacity and high resistance, in various computer systems but it has disadvantages which are the limited lifespan and imbalanced operation latency. Therefore, many page replacement policies have been studied to overcome the disadvantages of NAND flash memories. Although it is clear that these policies reflect execution characteristics of various environments and applications, researches on the foundation-policy decision for disk buffer management are insufficient. Thus, in this paper, we propose a foundation-policy recommendation model, called FRM for effectively utilizing NAND flash memories. FRM proposes a suitable page replacement policy by classifying and analyzing characteristics of workloads through machine learning. As an implementation case, we introduce FRM with a disk buffer management policy and in experiment results, prediction accuracy and weighted average of FRM shows 92.85% and 88.97%, by training dataset and validation dataset for foundation disk buffer management policy, respectively.

• Geomechanics and Engineering
• /
• v.34 no.3
• /
• pp.267-284
• /
• 2023

Landslides are one of the most dangerous phenomena and natural disasters. Landslides cause many human and financial losses in most parts of the world, especially in mountainous areas. Due to the climatic conditions and topography, people in the northern and western regions of Iran live with the risk of landslides. One of the measures that can effectively reduce the possible risks of landslides and their crisis management is to identify potential areas prone to landslides through multi-criteria modeling approach. This research aims to model landslide potential area in the Oshvand watershed using a support vector machine algorithm. For this purpose, evidence maps of seven effective factors in the occurrence of landslides namely slope, slope direction, height, distance from the fault, the density of waterways, rainfall, and geology, were prepared. The maps were generated and weighted using the continuous fuzzification method and logistic functions, resulting values in zero and one range as weights. The weighted maps were then combined using the support vector machine algorithm. For the training and testing of the machine, 81 slippery ground points and 81 non-sliding points were used. Modeling procedure was done using four linear, polynomial, Gaussian, and sigmoid kernels. The efficiency of each model was compared using the area under the receiver operating characteristic curve; the root means square error, and the correlation coefficient . Finally, the landslide potential model that was obtained using Gaussian's kernel was selected as the best one for susceptibility of landslides in the Oshvand watershed.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.23 no.3
• /
• pp.29-47
• /
• 2024

Many domestic and foreign studies derive factors that significantly affect the use of shared E-scooters based on performance data, but few studies have been conducted with comparative analysis models using predictive power, applying them to other regions. Therefore, by clearly establishing detailed influencing factors and scope in Gwangjin-gu and Gangnam-gu by using domestic shared E-scooter performance data, this study enhances predictive power, and the Geographically Weighted Regression model is derived through spatial autocorrelation verification. Based on the results, the direction of a construction model created from regional differences was presented, and major implications from the user's perspective are derived based on the difference between actual use and the model's prediction.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.127-137
• /
• 2019

Recidivism prediction has been a subject of constant research by experts since the early 1970s. But it has become more important as committed crimes by recidivist steadily increase. Especially, in the 1990s, after the US and Canada adopted the 'Recidivism Risk Assessment Report' as a decisive criterion during trial and parole screening, research on recidivism prediction became more active. And in the same period, empirical studies on 'Recidivism Factors' were started even at Korea. Even though most recidivism prediction studies have so far focused on factors of recidivism or the accuracy of recidivism prediction, it is important to minimize the prediction misclassification cost, because recidivism prediction has an asymmetric error cost structure. In general, the cost of misrecognizing people who do not cause recidivism to cause recidivism is lower than the cost of incorrectly classifying people who would cause recidivism. Because the former increases only the additional monitoring costs, while the latter increases the amount of social, and economic costs. Therefore, in this paper, we propose an XGBoost(eXtream Gradient Boosting; XGB) based recidivism prediction model considering asymmetric error cost. In the first step of the model, XGB, being recognized as high performance ensemble method in the field of data mining, was applied. And the results of XGB were compared with various prediction models such as LOGIT(logistic regression analysis), DT(decision trees), ANN(artificial neural networks), and SVM(support vector machines). In the next step, the threshold is optimized to minimize the total misclassification cost, which is the weighted average of FNE(False Negative Error) and FPE(False Positive Error). To verify the usefulness of the model, the model was applied to a real recidivism prediction dataset. As a result, it was confirmed that the XGB model not only showed better prediction accuracy than other prediction models but also reduced the cost of misclassification most effectively.