• Korean Journal of Cognitive Science
• /
• v.21 no.4
• /
• pp.603-625
• /
• 2010

Neural representation of a visual object is distributed across visual cortex and overlapped with those of many other objects. Thus repeating an object facilitates the recognition of the object while it impairs the recognition of other objects. These effects are called repetition priming and antipriming, respectively. Two experiments investigated a new phenomenon of repetition antipriming, in which a repeated object itself is antiprimed. The learning stage presented object pictures which were degraded at various levels. Participants determined how recognizable each object was. Then, the test stage presented the intact version of the object pictures and made participants to perform a categorization task. Both Experiment 1 and 2 found that the processing of the objects that had been recognized were facilitated (repetition priming) while the processing of the objects that had been perceptually ambiguous were impaired (repetition antipriming). These findings suggest that experiencing a perceptually ambiguous object might enhance the connection between feature-level representations and multiple object-level representations, which impairs the subsequent recognition of the repeated object.

• KIPS Transactions on Computer and Communication Systems
• /
• v.9 no.1
• /
• pp.9-16
• /
• 2020

The k-anonymity scheme has been widely used to protect private information when Big Data are distributed to a third party for research purposes. When the scheme is applied, an optimal k value determination is one of difficult problems to be resolved because many factors should be considered. Currently, the determination has been done almost manually by human experts with their intuition. This leads to degrade performance of the anonymization, and it takes much time and cost for them to do a task. To overcome this problem, a simple idea has been proposed that is based on machine learning. This paper describes implementations and experiments to realize the proposed idea. In thi work, a deep neural network (DNN) is implemented using tensorflow libraries, and it is trained and tested using input dataset. The experiment results show that a trend of training errors follows a typical pattern in DNN, but for validation errors, our model represents a different pattern from one shown in typical training process. The advantage of the proposed approach is that it can reduce time and cost for experts to determine k value because it can be done semi-automatically.

• Korean Journal of Applied Biomechanics
• /
• v.24 no.3
• /
• pp.287-293
• /
• 2014

The purpose of this study was to investigate the effects of Hoehn-Yahr scale on the activation of lower-extremity muscles during walking. Electromyography (EMG) analysis was carried out on 36 patients with Parkinson's disease in the off phase of the medication cycle. We recorded EMG signals of the tibialis anterior (TA), medial gastrocnemius (MG), lateral gastrocnemius (LG), soleus (SOL), rectus femoris (RF), vastus lateralis (VL), semitendinosus (ST) and biceps femoris (BF) using Noraxon 16 channels EMG system during walking at preferred speed. Rectified EMG signals were normalized to reference voluntary contractions (RVC) over a gait cycle at the preferred speed, allowing for an assessment of how the activity was distributed over the gait cycle. Compared to the H & Y Scale 1, H & Y Scale 3 exhibited greater activation of the vastus lateralis during mid-stance and greater activation of the medial gastrocnemius during terminal swing. Compared to the H & Y Scale 1, H & Y Scale 2 and 3 exhibited less activation of the tibialis anterior during initial swing. We conclude that the more Hoen & Yahr Scale increase, the more abnormal lower-extremity muscles activation.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• v.24 no.3
• /
• pp.371-382
• /
• 2014

Objectives: This study was undertaken in order to examine how musculoskeletal disorder(MSD) symptoms were affected by particular factors and then to explore control strategies to prevent MSDs in general hospital nurses. Materials: This, as part of a large study, was conducted using a set of information on literature review, questionnaire survey and focus group interview. It obtained prevalence and factors of MSD symptoms and examined how MSD symptoms were distributed and affected by the factors in nurses working at 15 general hospitals across Korea. The factors were personal factors, work organization, nursing tasks, physical factors and psychosocial factors. Results: A total of 501 nurses were determined as subjects. The highest MSD symptom prevalence was 61% for the shoulder, among body parts, followed by leg/feet(55%), low back(51%), neck(42%), wrist(38%), and elbow(21%). Prevalence for the whole body was 80%. Odds ratios ranged from 0.4 to 22.4 in logistic regression analyses. The symptoms were significantly attributed to factor variables such as body mass index, current health status, daily work time, nursing task, pooled-physical factors, ergonomic factors, work load, interpersonal conflict, and job insecurity. Conclusions: Two or more factor variables were significant, depending on body part, for MSD systems in the general hospital nurses. It was noticeable that physical factors, such as pooled-physical factors, ergonomic factors or work load, were selectively significant for MSD symptoms in all body parts, indicating that such information should be used for prevention of MSDs in the hospital sector.

• Journal of the Korea Society for Simulation
• /
• v.22 no.1
• /
• pp.77-86
• /
• 2013

Cloud computing is one of the distributed computing environments and utilizes several computing resources. Cloud environment uses a virtual machine to process a requested job. To balance a workload and process a job rapidly, cloud environment uses a provisioning technique and assigns a task with a status of virtual machine. However, a scheduling method for cloud computing requires a definition of virtual machine availabilities, which have an obscure meaning. In this paper, we propose Fuzzy logic driven Virtual machine Provisioning scheduling using Resource Evaluation(FVPRE). FVPRE analyzes a state of every virtual machine and actualizes a value of resource availability. Thus FVPRE provides an efficient provisioning scheduling with a precise evaluation of resource availability. FVPRE shows a high throughput and utilization for job processing on cloud environments.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.27 no.6
• /
• pp.679-688
• /
• 2009

Worldwide exploring and research for manganese nodules, as new energy resource, distributed on the deep seabed have progressed recently. Korea Ocean Research & Development Institute(KORDI) is a central organization to exploit the manganese nodules in the Pacific Ocean with 5,000m depth. Precise exploration is required for estimating amount of recoverable deposit, and this task could be accomplished by processing digital image processing techniques to the images taken by underwater camera system. Image processing and analysis provide information about characteristics of distribution of the manganese nodules. This study proposed effective methods to remove vignetting effect to improve image quality and to extract information. The results show more reliable information could be obtained by removing the vignetting and feasibility of utilizing image processing techniques for exploring the manganese nodules.

• The KIPS Transactions:PartB
• /
• v.10B no.1
• /
• pp.1-12
• /
• 2003

In this paper, we Introduce the Unreal Tournament (UT) game and the Gamebots system. The former it a well-known 3D first-person action game and the latter is an intelligent agent research testbed based on UT And then we explain the design and implementation of KGBot, which is an intelligent non-player character deploying effectively within the 3D virtual environment provided by UT and the Gamebots system. KGBot is a bot client within the Gamebots System. KGBot accomplishes its own task to find out and dominate several domination points pro-located on the complex surface map of 3D virtual environment KGBot adopts UM-PRS as its control engine, which is a general BDI agent architecture. KGBot contains a hierarchical knowledge base representing its complex behaviors in multiple layers. In this paper, we explain details of KGBot's Intelligent behaviors, tuck af locating the hidden domination points by exploring the unknown world effectively. constructing a path map by collecting the waypoints and paths distributed over the world, and finding an optimal path to certain destination based on this path graph. Finally we analyze the performance of KGBot exploring strategy and control engine through some experiments on different 3D maps.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.5 no.9
• /
• pp.1492-1512
• /
• 2011

Due to multiple hops, mobility and time-varying channel, supporting delay sensitive real-time traffic in wireless local area network-based (WLAN) mesh networks is a challenging task. In particular for real-time traffic subject to medium access control (MAC) layer control overhead, such as preamble, carrier sense waiting time and the random backoff period, the performance of real-time flows will be degraded greatly. In order to support real-time traffic, an efficient adaptive packet scheduling (APS) scheme is proposed, which aims to improve the system performance by guaranteeing inter-class, intra-class service differentiation and adaptively adjusting the packet length. APS classifies incoming packets by the IEEE 802.11e access class and then queued into a suitable buffer queue. APS employs strict priority service discipline for resource allocation among different service classes to achieve inter-class fairness. By estimating the received signal to interference plus noise ratio (SINR) per bit and current link condition, APS is able to calculate the optimized packet length with bi-dimensional markov MAC model to improve system performance. To achieve the fairness of intra-class, APS also takes maximum tolerable packet delay, transmission requests, and average allocation transmission into consideration to allocate transmission opportunity to the corresponding traffic. Detailed simulation results and comparison with IEEE 802.11e enhanced distributed channel access (EDCA) scheme show that the proposed APS scheme is able to effectively provide inter-class and intra-class differentiate services and improve QoS for real-time traffic in terms of throughput, end-to-end delay, packet loss rate and fairness.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.67 no.4
• /
• pp.183-190
• /
• 2018

Predicting accurate electricity prices is an important task in the electricity trading market. To address the electricity price forecasting problem, various approaches have been proposed so far and it is known that linear regression-based approaches are the best. However, the use of such linear regression-based methods is limited due to low accuracy and performance. In traditional linear regression methods, it is not practical to find a nonlinear regression model that explains the training data well. If the training data is complex (i.e., small-sized individual data and large-sized features), it is difficult to find the polynomial function with n terms as the model that fits to the training data. On the other hand, as a linear regression model approximating a nonlinear regression model is used, the accuracy of the model drops considerably because it does not accurately reflect the characteristics of the training data. To cope with this problem, we propose a new electricity price forecasting method that divides the entire dataset to multiple split datasets and find the best linear regression models, each of which is the optimal model in each dataset. Meanwhile, to improve the performance of the proposed method, we modify the proposed localized linear regression method in the map and reduce way that is a framework for parallel processing data stored in a Hadoop distributed file system. Our experimental results show that the proposed model outperforms the existing linear regression model. Specifically, the accuracy of the proposed method is improved by 45% and the performance is faster 5 times than the existing linear regression-based model.

• Journal of Information Technology Services
• /
• v.21 no.4
• /
• pp.63-74
• /
• 2022

Globally researchers at medical institutions are actively sharing COHORT data of patients to develop vaccines and treatments to overcome the COVID-19 crisis. OMOP-CDM, a common data model that efficiently shares medical data research independently operated by individual medical institutions has patient personal information (e.g. PII, PHI). Although PII and PHI are managed and shared indistinguishably through de-identification or anonymization in medical institutions they could not be guaranteed at 100% by complete de-identification and anonymization. For this reason the security of the OMOP-CDM database is important but there is no detailed and specific OMOP-CDM security inspection tool so risk mitigation measures are being taken with a general security inspection tool. This study intends to study and present a model for implementing a tool to check the security vulnerability of OMOP-CDM by analyzing the security guidelines for the US database and security controls of the personal information protection of the NIST. Additionally it intends to verify the implementation feasibility by real field demonstration in an actual 3 hospitals environment. As a result of checking the security status of the test server and the CDM database of the three hospitals in operation, most of the database audit and encryption functions were found to be insufficient. Based on these inspection results it was applied to the optimization study of the complex and time-consuming CDM CSF developed in the "Development of Security Framework Required for CDM-based Distributed Research" task of the Korea Health Industry Promotion Agency. According to several recent newspaper articles, Ramsomware attacks on financially large hospitals are intensifying. Organizations that are currently operating or will operate CDM databases need to install database audits(proofing) and encryption (data protection) that are not provided by the OMOP-CDM database template to prevent attackers from compromising.