• 한국경영과학회지
• /
• 제39권4호
• /
• pp.51-69
• /
• 2014

Times to multiple events (TMEs) are a major data type in large-scale business and medical data. Despite its importance, the analysis of TME data has not been well studied because of the analysis difficulty from censoring of observation. To address this difficulty, we have developed a Bayesian-based multivariate survival analysis method, which can successfully estimate the joint probability density of survival times. In this work, we extended this method for the analysis of precedence, dependency and causality among multiple events. We applied this method to the electronic health records of 2,111 patients in a children's hospital in the US and the proposed analysis successfully shows the relation between times to two types of hospital visits for different medical issues. The overall result implies the usefulness of the multivariate survival analysis method in large-scale big data in a variety of areas including marketing, human resources, and e-commerce. Lastly, we suggest our future research directions based multivariate survival analysis method.

• 식물분류학회지
• /
• 제49권2호
• /
• pp.127-139
• /
• 2019

The purported four hybrid origins of Asplenium in Korea were tested based on morphological, cytological and DNA sequence data. Asplenium castaneo-viride, A. ${\times}$ uiryeongse, A. ${\times}$ montanus, and A. ${\times}$ kitazawae share several morphological characteristics with the Asian walking fern A. ruprechtii and related taxa as parents and show a sympatric distribution with the putative parents, raising the possibility of hybrid origins: A. castaneo-viride (A. ruprechtii and A. incisum), A. ${\times}$ uiryeongse (A. ruprechtii and A. pekinense), A. ${\times}$ montanus (A. ruprechtii, A. trichomanes, and A. incisum), and A. ${\times}$ kitazawae (A. ruprechtii and A. sarelii). We investigated flow cytometry and chloroplast DNA sequence data (rbcL, rps4-trnS, and rps4-trnS intergenic spacer) to clarify the hybridization and origin of each hybrid. In the flow cytometry analyses, A. ruprechtii shows diploid (2x) only, whereas A. castaneo-viride (3x, 4x), A. ${\times}$ uiryeongse (3x), A. ${\times}$ montanus (3x, 4x), and A. ${\times}$ kitazawae (2x, 4x) exhibit polyploidy, suggesting hybrid events along speciation. The rbcL and rps4-trnS and rps4-trnS intergenic spacer data suggest that A. ruprechtii is one the maternal ancestors of all four hybrids. In addition, the rps4-trnS and rps4-trnS intergenic spacer data indicate that A. incisum is also the maternal ancestor of A. ${\times}$ kitazawae and A. ${\times}$ montanus, proposing multiple hybridization events for these two hybrids. In A. ${\times}$ montanus, morphological features such as the leaf forms and sympatric distributions of the species also support the multimaternal hypothesis, but the morphological features of A. ${\times}$ kitazawae must be examined with consideration of hybrid events. To clarify the complex hybrid evolutionary lineages of the four Asplenium hybrids, further research with taxon sampling and molecular markers should be conducted.

• Structural Engineering and Mechanics
• /
• 제30권3호
• /
• pp.369-382
• /
• 2008

An oscillator of two lumped masses linked through a vertical spring moves forward in the horizontal direction, initially at a certain height, over a horizontal Euler beam and descends on it due to its own weight. Vibration of the beam and the oscillator is excited at the onset of the ensuing impact. The impact produced by the descending oscillator is assumed to be either perfectly elastic or perfectly plastic. If the impact is perfectly elastic, the oscillator bounces off and hits the beam a number of times as it moves forward in the longitudinal direction of the beam, exchanging its dynamics with that of the beam. If the impact is perfectly plastic, the oscillator (initially) sticks to the beam after its first impact and then may separate and reattach to the beam as it moves along the beam. Further events of separation and reattachment may follow. This interesting and seemingly simple dynamic problem actually displays rather complicated dynamic behaviour and has never been studied in the past. It is found through simulated numerical examples that multiple events of separation and impact can take place for both perfectly elastic impact and perfectly plastic impact (though more of these in the case of perfectly elastic impact) and the dynamic response of the oscillator and the beam looks noisy when there is an event of impact because impact excites higher-frequency components. For the perfectly plastic impact, the oscillator can experience multiple events of consecutive separation from the beam and subsequent reattachment to it.

• Nuclear Engineering and Technology
• /
• 제43권3호
• /
• pp.271-278
• /
• 2011

When performing fire PSA in a nuclear power plant, an event mapping method, using an internal event PSA model, is widely used to reduce the resources used by fire PSA model development. Feasible initiating events and component failure events due to fire are identified to transform the fault tree (FT) for an internal event PSA into one for a fire PSA using the event mapping method. A surrogate event or damage term method is used to condition the FT of the internal PSA. The surrogate event or the damage term plays the role of flagging whether the system/component in a fire compartment is damaged or not, depending on the fire being initiated from a specified compartment. These methods usually require explicit states of all compartments to be modeled in a fire area. Fire event scenarios, when using explicit identification, such as surrogate or damage terms, have two problems: (1) there is no consideration of multiple fire propagation beyond a single propagation to an adjacent compartment, and (2) there is no consideration of simultaneous fire propagations in which an initiating fire event is propagated to multiple paths simultaneously. The present paper suggests a fire propagation equation to identify all possible fire event scenarios for an explicitly treated fire event scenario in the fire PSA. Also, a method for separating fire events was developed to make all fire events a set of mutually exclusive events, which can facilitate arithmetic summation in fire risk quantification. A simple example is given to confirm the applicability of the present method for a $2{\times}3$ rectangular fire area. Also, a feasible asymptotic approach is discussed to reduce the computational burden for fire risk quantification.

• 천문학회보
• /
• 제37권2호
• /
• pp.110.1-110.1
• /
• 2012

We studied the source regions of 12 solar energetic particle (SEP) events seen between 2010 August and 2012 January at STEREO-A, B and ACE, when the two STEREO spacecraft were separated by about $180^{\circ}$. All events were associated with strong flares (C1 - X6) and fast coronal mass ejections (CMEs) accompanied by type II radio bursts. We have determined the arrival times of the SEP events at the three spacecraft. EUV waves observed in $195{\AA}$ and $193{\AA}$ channels of STEREO and SDO/AIA are tracked across the Sun and the arrival time of the EUV wave at the photospheric source of open field lines extending to the spacecraft connection points at 2.5 Rsun estimated. We found 7 events with flux enhancements in all spacecraft and 4 in two spacecraft. Most events came from a single source. The results show that magnetic field connections between source regions and the spacecraft play an important role in abrupt flux enhancements. In the most cases, EUV waves at the Sun are associated with a wide longitudinal spread of the SEPs.

• 한국철도학회:학술대회논문집
• /
• 한국철도학회 2001년도 추계학술대회 논문집
• /
• pp.198-204
• /
• 2001

Train crashes involve complex interaction between deformable bodies in multiple collisions. The purpose of this study is to suggest the effective analytical procedure using hybrid model for the crashworthiness of motorized trailer of high speed train. The hybrid approach, with very short modeling times and reduced computation times to extract the global behaviour and to perform a pre-optimization of the considered structure. Firstly, various types of crash events are investigated and the conditions for numerical simulation are defined. In this paper, the structural crashwonhiness of Korean High Speed Train trailer was examined through FE analysis. Crash analyses on energy absorbing part and safety zone were carried out to determine each section force. Rollover analysis was performed to observe the amount of intrusion in the passenger's area in case of rollover accident.

• 한국철도학회:학술대회논문집
• /
• 한국철도학회 2002년도 추계학술대회 논문집(I)
• /
• pp.80-85
• /
• 2002

Train crashes involve complex interaction between deformable bodies in multiple collisions. The purpose of this study is to suggest the effective analytical procedure using hybrid model fur the crashworthiness of motorized trailer of high speed train. The hybrid approach, with very short modeling times and reduced computation times to extract the global behaviour and to perform a pre-optimization of the considered structure. Firstly, various types of crash events are investigated and the conditions for numerical simulation are defined. In this paper, the structural stiffness and crashworthiness of Korean High Speed Train trailer was examined through FE analysis. Crash analyses on energy absorbing part and safety zone were carried out to determine each section force.

• 대한임베디드공학회논문지
• /
• 제14권5호
• /
• pp.277-286
• /
• 2019

In this paper, we propose an efficient hyperplane generation technique to classify human activity from combination of events and sequence information obtained from multiple-event sensors. By generating hyperplane efficiently, our machine learning algorithm classify with less memory and run time than the LSVM (Linear Support Vector Machine) for embedded system. Because the fact that light weight and high speed algorithm is one of the most critical issue in the IoT, the study can be applied to smart home to predict human activity and provide related services. Our approach is based on reducing numbers of hyperplanes and utilizing robust string comparing algorithm. The proposed method results in reduction of memory consumption compared to the conventional ML (Machine Learning) algorithms; 252 times to LSVM and 34,033 times to LSTM (Long Short-Term Memory), although accuracy is decreased slightly. Thus our method showed outstanding performance on accuracy per hyperplane; 240 times to LSVM and 30,520 times to LSTM. The binarized image is then divided into groups, where each groups are converted to binary number, in order to reduce the number of comparison done in runtime process. The binary numbers are then converted to string. The test data is evaluated by converting to string and measuring similarity between hyperplanes using Levenshtein algorithm, which is a robust dynamic string comparing algorithm. This technique reduces runtime and enables the proposed algorithm to become 27% faster than LSVM, and 90% faster than LSTM.

• 천문학회보
• /
• 제39권1호
• /
• pp.75.2-75.2
• /
• 2014

We have developed a set of daily solar flare peak flux forecast models using the multiple linear regression (MLR), the auto regression (AR), and artificial neural network (ANN) methods. We consider input parameters as solar activity data from January 1996 to December 2013 such as sunspot area, X-ray flare peak flux, weighted total flux $T_F=1{\times}F_C+10{\times}F_M+100{\times}F_X$ of previous day, mean flare rates of a given McIntosh sunspot group (Zpc), and a Mount Wilson magnetic classification. We compute the hitting rate that is defined as the fraction of the events whose absolute differences between the observed and predicted flare fluxes in a logarithm scale are ${\leq}$ 0.5. The best three parameters related to the observed flare peak flux are as follows: weighted total flare flux of previous day (r=0.5), Mount Wilson magnetic classification (r=0.33), and McIntosh sunspot group (r=0.3). The hitting rates of flares stronger than the M5 class, which is regarded to be significant for space weather forecast, are as follows: 30% for the auto regression method and 69% for the neural network method.

• 천문학회보
• /
• 제37권1호
• /
• pp.95.2-95.2
• /
• 2012

We have compared the geoeffective parameters of halo coronal mass ejections (CMEs) to predict geomagnetic storms. For this we consider 50 front-side full halo CMEs whose asymmetric cone model parameters and earthward direction parameter were available. For each CME we use its projected velocity (Vp), radial velocity (Vr), angle between cone axis and sky plane (${\gamma}$) from the cone model, earthward direction parameter (D), source longitude (L), and magnetic field orientation (M) of the CME source region. We make a simple and multiple linear regression analysis to find out the relationship between CME parameters and Dst index. Major results are as follows. (1) $Vr{\times}{\gamma}$ has a higher correlation coefficient (cc = 0.70) with the Dst index than the others. When we make a multiple regression of Dst and two parameters ($Vr{\times}{\gamma}$, D), the correlation coefficient increases from 0.70 to 0.77. (2) Correlation coefficients between Dst index and $Vr{\times}{\gamma}$ have different values depending on M and L. (3) Super geomagnetic storms (Dst ${\leq}$ -200 nT) only appear in the western and southward events. Our results demonstrate that not only the cone model parameters together with the earthward direction parameter improve the relationship between CME parameters and Dst index but also the source longitude and its magnetic field orientation play a significant role in predicting geomagnetic storms.