• 한국시뮬레이션학회:학술대회논문집
• /
• 한국시뮬레이션학회 1993년도 제3회 정기총회 및 추계학술발표회
• /
• pp.2-2
• /
• 1993

Inference for discrete event simulations usually relies on either independent replications or, if each simulation run is expensive, the method of batch means applied to a single replications. We present a new method, threshold bootstrap, which equals or exceeds the performance of independent replications or batch means. The method works by resampling runs of data created when a stationary time series crosses a threshold level, such as the sample mean of series. Computational results show that the threshold bootstrap matches or exceeds the performance of these alternative methods in estimating the standard deviation of the sample mean and producing valid confidence intervals.

• Journal of Biosystems Engineering
• /
• 제24권2호
• /
• pp.145-152
• /
• 1999

Methods of discharging each graded agricultural product are divided into two according to the type of feeding. One is based on feeding objects using a series of specially designed holders mounted with an equal interval. The other is randomly feeding objects while being isolated without a specific interval. In this paper, a real time simultaneous discharge algorithm, which is good for objects randomly fed individually and for multi-objects located along a series of discharge buckets, was developed. And the developed algorithm was implemented to the controller and the performance was verified using the system developed for dried mushrooms. The discharge system used for the experiment was composed of a variable speed conveyor, a series of double channel bucket mounted along both sides of the conveyor, and a series of air nozzles and optic sensors. Developed algorithm worked perfectly and could be directly used for automatic discharge system for randomly feeding agricultural products.

• 한국지하수토양환경학회:학술대회논문집
• /
• 한국지하수토양환경학회 2001년도 추계학술발표회
• /
• pp.135-138
• /
• 2001

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권8호
• /
• pp.4108-4122
• /
• 2019

Workflow process mining is becoming a more and more valuable activity in workflow-supported enterprises, and through which it is possible to achieve the high levels of qualitative business goals in terms of improving the effectiveness and efficiency of the workflow-supported information systems, increasing their operational performances, reducing their completion times with minimizing redundancy times, and saving their managerial costs. One of the critical challenges in the workflow process mining activity is to devise a reasonable approach to discover and recognize the bottleneck points of workflow process models from their enactment event histories. We have intuitively realized the fact that the iterative process pattern of redo-activities ought to have the high possibility of becoming a bottleneck point of a workflow process model. Hence, we, in this paper, propose an algorithmic approach and its implementation to discover the redo-activities and their performers' involvements patterns from workflow process enactment event logs. Additionally, we carry out a series of experimental analyses by applying the implemented algorithm to four datasets of workflow process enactment event logs released from the BPI Challenges. Finally, those discovered redo-activities and their performers' involvements patterns are visualized in a graphical form of information control nets as well as a tabular form of the involvement percentages, respectively.

• 한국음향학회지
• /
• 제35권2호
• /
• pp.143-148
• /
• 2016

깊은 신경망은 데이터의 특성을 효과적으로 나타낼 수 있는 방법으로 최근 많은 응용 분야에서 활용되고 있다. 하지만, 제한적인 양의 데이터베이스는 깊은 신경망을 훈련하는 과정에서 과적합 문제를 야기할 수 있다. 본 논문에서는 풍부한 양의 음성 혹은 음악 데이터를 이용한 전이학습을 통해 제한적인 양의 사운드 이벤트에 대한 깊은 신경망을 효과적으로 훈련하는 방법을 제안한다. 일련의 실험을 통해 제안하는 방법이 적은 양의 사운드 이벤트 데이터만으로 훈련된 깊은 신경망에 비해 현저한 성능 향상이 있음을 확인하였다.

• 한국컴퓨터정보학회논문지
• /
• 제27권8호
• /
• pp.169-176
• /
• 2022

5G와 인공지능, 자율 주행 차량 시스템에 관한 기술의 지속적인 발전에 힘입어 V2X를 비롯한 C-ITS 환경에 대한 표준과 서비스가 연구되고 있다. V2V 환경에서 차량 시스템에서 수집 및 생성되는 데이터를 기반 차량간 데이터를 교환하기 위한 표준으로 BSM(basic safety message) 이 채택되었다. 본 논문에서는 BSM 메시지를 안전하게 저장하고, Event-Sourcing과 CQRS 패턴을 사용해서 저장된 메시지를 효과적으로 확인할 수 있는 프레임워크를 제안한다. 제안된 프레임워크는 해시 함수를 사용해서 BSM 메시지를 안전하게 저장 및 관리할 수 있으며, 시계열을 기반으로 저장된 BSM 데이터를 실시간으로 확인할 수 있으며 상태 재현이 가능한 장점이 있다.

• 인터넷정보학회논문지
• /
• 제20권6호
• /
• pp.137-142
• /
• 2019

본 논문은 이상 상황을 탐지하고 모니터링하는 다양한 서비스가 존재한다. 하지만 대부분의 서비스는 화재, 가스누출에 초점을 맞추어 진행되고 있으며, 독거노인과 중증장애인들의 사망 혹은 심정지 등 위급상황에 대하여 사전 예방 및 위급상황 대응이 불가능하다. 본 연구에서는 여러 생체신호 중 가장 위중하다고 판단되는 심박 신호의 이상 상태를 탐지하기 위하여 인공지능 모델을 설계하는 과정에서 적합한 데이터 변형과 모델을 비교한다. 세부적으로는 오픈 의료 데이터 PhysioNet의 MIT-BIH Arrhythmia Database를 이용하여 심전도(ECG) 데이터를 수집하고, 수집한 데이터를 각각 다른 방법으로 데이터를 변형한 후 학습하여 기본 심전도 데이터를 이용해 학습한 인공지능 모델과 비교한다.

• 대한의용생체공학회:학술대회논문집
• /
• 대한의용생체공학회 1997년도 춘계학술대회
• /
• pp.275-278
• /
• 1997

Standard methods estimating the power spectral density(PSD) from an irregularly sampled cardiac event series require deriving a new evenly-spaced signal applicable to those methods. To avoid that requirement, in this study, the power spectrum of heart rate variability was estimated by Lomb-Scargle's algorithm, which is a means of obtaining PSD estimates directly from irregularly sampled timeseries observed in astronomy. To assess the performance of Lomb-Scargle algorithm in the power spectral analysis of heart rate variability, it was applied to various cardiac event series derived through integral pulse frequency modulation model(IPFM) simulation and from real ECG signals, and the resultant power spectra was compared with those obtained by a conventional method based on the FFT. In result, it is concluded that Lomb-Scargle's periodogram is very effective in the power spectral analysis of heart rate variability, especially in the presence of arrhythmia and/or dropouts of cardiac events.

• Journal of the korean society of oceanography
• /
• 제39권4호
• /
• pp.207-211
• /
• 2004

We used long time series of hydrographic and biological variables to examine the ecosystem consequences of a rare, anomalous event in the south sea of Korea. The highest zooplankton biomass in 36 years of sampling occurred in April 1997. Zooplankton biomass exceeded 2 times than the long-term mean at 35% of the stations. Copepod abundance was low in April and June and also failed to show a seasonal peak in 1997. Mackerel (Scomber japonicus) catches were very low in spring 1997 and 1999, in spite of a positive correlation between zooplankton biomass and mackerel catches at lags of 0, 12 and 24 months. It was discussed that a high zooplankton biomass with low copepod abundance in April 1997 resulted from unusual high temperature and salps abundance. Water temperatures were ca. $2^{\circ}C$ higher than the long-term mean at the surface. Salps and doliolids (thaliaceans), especially the warm-water species Doliolum nationalis, dominated the zooplankton. An unusual incursion of the Tsushima Warm Current may have transported the thaliaceans into the area and/or produced favorable conditions for a bloom. This study suggested that taxonomic composition of zooplankton was important to decide mackerel catches.

• 천문학회보
• /
• 제40권2호
• /
• pp.55.4-56
• /
• 2015

The Hilbert-Huang Transform (HHT) is composed of the Empirical Mode Decomposition (EMD) and the Hilbert Spectral Analysis (HSA). The EMD decomposes any time series data into a small number of components called the Intrinsic Mode Functions (IMFs), compared to the Discrete Fourier Transform which decomposes a data into a large number of harmonic functions. Each IMF has varying amplitude and frequency with respect to time, which can be obtained by HSA. The time resolution of the modes in HHT is the same as that of the given time series, while in the Wavelet Transform, Constant Q Transform and Short-Time Fourier Transform, there is a tradeoff between the resolutions in frequency and time. Based on the time-dependent amplitudes of IMFs, we develop an Event Trigger Generator and demonstrate its efficiency by applying it to gravitational-wave data.