• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.11
• /
• pp.449-456
• /
• 2021

An intrusion detection system is a technology that detects abnormal behaviors that violate security, and detects abnormal operations and prevents system attacks. Existing intrusion detection systems have been designed using statistical analysis or anomaly detection techniques for traffic patterns, but modern systems generate a variety of traffic different from existing systems due to rapidly growing technologies, so the existing methods have limitations. In order to overcome this limitation, study on intrusion detection methods applying various machine learning techniques is being actively conducted. In this study, a comparative study was conducted on data preprocessing techniques that can improve the accuracy of anomaly detection using NGIDS-DS (Next Generation IDS Database) generated by simulation equipment for traffic in various network environments. Padding and sliding window were used as data preprocessing, and an oversampling technique with Adversarial Auto-Encoder (AAE) was applied to solve the problem of imbalance between the normal data rate and the abnormal data rate. In addition, the performance improvement of detection accuracy was confirmed by using Skip-gram among the Word2Vec techniques that can extract feature vectors of preprocessed sequence data. PCA-SVM and GRU were used as models for comparative experiments, and the experimental results showed better performance when sliding window, skip-gram, AAE, and GRU were applied.

• Reproductive and Developmental Biology
• /
• v.36 no.1
• /
• pp.65-70
• /
• 2012

For reconstituting genetic resource(Korean Native Chicken: KNC) with grem-line chimeric chicken made with cryopreserved biastdermal cells, the experiments were carried out to optimize cryopreservating conditions. Stage X biastdemal cells were collected from KNC embryos and dissociated. Cells were susupended in medium containing cyopretectant and fetal bovine serum(FBS), and distributed into plastic ampules. Cell susupensions were seeded to induce ice formation at $-7^{\circ}C$ to $-35^{\circ}C$ at in the experiments, the effect of modification of dissociation way, concentration of FBS and cell density on the vaibility of frezen-thawed cells were investigated by trypan blue exclusion. Then change the way of cell dissociation from pipetting to short time vortexing, viability of frozen-thawed cell tended to be increaced from 29 % to 52 %. Increase concentraition of FBS in frozen medium from 20 % to 80 % made viability of thawed cell from 28 % to 35 %. The viability of thawed cells were 33.9% frozen at 2 embryos/0.5 ml, and 43.6 % frozen at 20 embryos/0.5 ml. Furthermore, combination of three modifications make big improvement. The viability of frozen-thawed cell was 60 % for combinated method, and 41 % for general method. This result means the advance to practical cryoreservation of blastdermal cell of the KNC(Ogolgye breed).

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.3
• /
• pp.141-148
• /
• 2022

Recently, reinforcement learning combined with deep neural network technology has achieved remarkable success in various fields such as board games such as Go and chess, computer games such as Atari and StartCraft, and robot object manipulation tasks. However, such deep reinforcement learning describes states, actions, and policies in vector representation. Therefore, the existing deep reinforcement learning has some limitations in generality and interpretability of the learned policy, and it is difficult to effectively incorporate domain knowledge into policy learning. On the other hand, dNL-RRL, a new relational reinforcement learning framework proposed to solve these problems, uses a kind of vector representation for sensor input data and lower-level motion control as in the existing deep reinforcement learning. However, for states, actions, and learned policies, It uses a relational representation with logic predicates and rules. In this paper, we present dNL-RRL-based policy learning for transportation mobile robots in a manufacturing environment. In particular, this study proposes a effective method to utilize the prior domain knowledge of human experts to improve the efficiency of relational reinforcement learning. Through various experiments, we demonstrate the performance improvement of the relational reinforcement learning by using domain knowledge as proposed in this paper.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.3
• /
• pp.115-122
• /
• 2019

For Korean phoneme recognition, Hidden Markov-Gaussian Mixture model(HMM-GMM) or hybrid models which combine artificial neural network with HMM have been mainly used. However, current approach has limitations in that such models require force-aligned corpus training data that is manually annotated by experts. Recently, researchers used neural network based phoneme recognition model which combines recurrent neural network(RNN)-based structure with connectionist temporal classification(CTC) algorithm to overcome the problem of obtaining manually annotated training data. Yet, in terms of implementation, these RNN-based models have another difficulty in that the amount of data gets larger as the structure gets more sophisticated. This problem of large data size is particularly problematic in the Korean language, which lacks refined corpora. In this study, we introduce CTC algorithm that does not require force-alignment to create a Korean phoneme recognition model. Specifically, the phoneme recognition model is based on convolutional neural network(CNN) which requires relatively small amount of data and can be trained faster when compared to RNN based models. We present the results from two different experiments and a resulting best performing phoneme recognition model which distinguishes 49 Korean phonemes. The best performing phoneme recognition model combines CNN with 3hop Bidirectional LSTM with the final Phoneme Error Rate(PER) at 3.26. The PER is a considerable improvement compared to existing Korean phoneme recognition models that report PER ranging from 10 to 12.

• Korean Journal of Animal Reproduction
• /
• v.10 no.2
• /
• pp.192-203
• /
• 1986

Two experiments in this study were designed to compare the potential for in vitro capacitation and in vitro fertilization of ejaculated sperm among individual rabbit bucks. In experiment 1, for in vitro capacitation, the ejaculated sperm were preincubated in DM for 12 hr or 18 hr after HIS treatment, then 12 hr -or 18 hr- preincubated sperm were incubated with superovulated rabbit ova in a 5% CO2 incubator for 36 hr at 38$^{\circ}C$, and a part of cleaved ova was transferred to the recipient does for implantation of embryo. In experiment 2, effect of lysolecithin addition to preincubation medium on induction of accelerated in vitro capacitation and in vitro fertilization of individual rabbit sperm was studied. Experiment 1; 1. Percent acrosome reaction of sperm, noted after staining, after 12 hr or 18 hr preincubation ranged from 52.5 to 76.0% and from 67.5 to 90.0%, respectively and sperm motility index of these sperm ranged from 20.0 to 47.5 for 12 hr-preincubated sperm and from 15.0 to 37.5 for 18 hr- preincubated sperm. There was no a certain relation between percent acrosome reaction and sperm motility index. 2. In vitro fertilization rate (cleavage rate) of in vitro capacitated sperm varied widely among individual bucks, ranging from 0 to 47.8% for 12 hr - preincubated sperm and from 0 to 60.9% for 18 hr -prein- cubated sperm. Cleavage rate of 18 hr - preincubated sperm was higher and faster than that of 12 hr - preincubated sperm. 3. Eight of 44 in vitro fertilized embryos transferred into 6 recipients were implanted in 4 recipients (66.7%) up to day 15 and implnatation rate was 18.2%. Experiment 2; 1. The percent acrosome reaction of sperm before and after 4 hr preincubation in DM without lysolecithin varied significantly among individual bucks, ranging from 0.4 to 18.4% and from 1.7 to 37.4%, respectively and percent acrosome reaction of sperm at 30 min after addition of 60${\mu}$g/ml lysolecithin also was significantly different among bucks, ranging from 19.2 to 67.1%. 2. Effect of accelerated acrosome reaction following lysolecithin addition was more considerable in the individuals showed less percent acrosome reaction before and after 4 hr preincubation. Percentage of motile sperm and motility score showed a trendency towards a decrease with increase of preincubation time and time after lysolecithin addition. 3. In vitro fertilization rate (cleavage rate) at 24 hr postinesmination with pooled sperm were treated to 60 $\mu\textrm{g}$/ml lysolecithin for 30 min after 4 hr preincubation was 24.6%, a higher rate than 13.2% for control. While 80 $\mu\textrm{g}$/ml lysolecithin-added sperm showed a lower cleavage than control and 60$\mu\textrm{g}$/ml-added sperm at both 24 hr and 48 hr postinsemination. These results from 2 experiments suggest that more useful preincubation time for the in vitro capacitation of ejaculated rabbit sperm is 18 hr in DM after HIS treatment, although there is wide variation in vitro capacitation and in vitro fertilization rate among individual bucks, and lysolecithin addition to at least 4 hr - preincubated sperm in DM can result in almost same in vitro fertilization rate as that of 18 hr - preincubated sperm in the experiment 1.

• Journal of Broadcast Engineering
• /
• v.4 no.2
• /
• pp.103-118
• /
• 1999

This paper analyzes the performance of symbol timing recovery and equalizer in 8VSB digital terrestrial TV receiver under various multipath signals and proposes equalization techniques which improve indoor reception performance. Data segment sync is used for symbol timing recovery and timing offset is measured for echoes of various delays and amplitudes by using symbol timing detection filter whose pattern is +1. +1. -1. and -1. Measured timing offsets were below 10% for long echoes with more than 5 symbol delay and above 30% for short echoes with around 1 symbol delay. Indoor reception is always more challenging than outdoor reception due to lower signal strength. large and short multipaths. and moving interfering objects. So it is considered to use FSE (Fractionally Spaced Equalizer) which is very robust to timing offset and blind equalizer which can update equalizer tap coefficients even by information data. We compare the performance of conventional DFE (Decision Feedback Equalizer) and FSE-DFE using LMS algorithm and Stop and Go algorithm for the indoor reception. Experiments reveals FSE has excellent performance for large timing offset and Stop and Go algorithm shows good performance for Doppler shift. so we propose to use FSE-DFE structure with Stop and Go algorithm for the reliable indoor reception.

• Journal of the Korea Institute of Building Construction
• /
• v.15 no.3
• /
• pp.281-289
• /
• 2015

Carbonation of concrete causes reduction of pH and subsequently causes steel corrosion for reinforced concrete structure. However, for plain concrete structure or PC product, it can lead to a decrease in porosity, high density, improvement of concrete, shrinkage-compensation. Recently, based on this theory, research of $CO_2$ curing effect has been performed, but it was mainly focused on its effects on compressive strength using only ordinary portland cement. Researches on $CO_2$ curing effect for concrete containing $CO_2$ reactive materials such as ${\gamma}-C_2S$, MgO haven't been investigated. Therefore, this study has performed experiments under water-binder ratio 40%, and the replacement ratios of ${\gamma}-C_2S$ and MgO were 90%. Micro-chemical analysis, measurement of compressive strength according to admixtures and $CO_2$ curing were investigated. Results from this study revealed that higher strength was measured in case of $CO_2$ curing compared with none $CO_2$ curing for plain specimen indicating difference between 1.08 and 1.26 times, in case of ${\gamma}-C_2S$ 90, MgO 90 specimen, incorporating high volume replaced as much as 90%, it was proven that when applying $CO_2$ curing, higher strength which has difference between 14.56 and 45.7 times, and between 6.5 and 10.37 times was measured for each specimen compared to none $CO_2$ curing. Through micro-chemical analysis, massive amount of $CaCO_3$, $MgCO_3$ and decrease of porosity were appeared.

• Journal of Nuclear Fuel Cycle and Waste Technology(JNFCWT)
• /
• v.8 no.4
• /
• pp.311-318
• /
• 2010

U. S. NRC Regulation Guide 1.52 regulating ESF ACS in nuclear power plants has been revised to revision 3. To apply reduction of operability test time, allowance of alternative challenge agents for in-place leak test of HEPA filters, and upgrade of Methyl Iodide penetration acceptance criterion in activated carbon performance test suggested in Reg. Guide 1.52(Rev.3) on Yonggwang units 5 and 6 ESF ACSes, technical feasibility study was carried out with on-site experiments as well as experiments with a lab-scale model. It was confirmed that the moisture in the system returned to the level before the test in 1 or 4 days even though the moisture was removed during the operability test lasting more than 10 hours. Therefore, it is appropriate to perform monthly operability test in 15 minutes just long enough to check the operability of equipment. To change challenge material for in-place HEPA filter leak test, size of aerosol, production rate, and leak detection capability were compared for DOP and PAO. It was concluded that PAO can be substituted for DOP in nuclear power plants. The upgrade of Methyl Iodide penetration acceptance criterion from 0.175 % to 0.5 % in active carbon filter bed deeper than 4 inches was to conform to the change of activated carbon performance test method to ASTM D3803(1989). It was confirmed that Methyl Iodide penetration acceptance criterion of 0.5 % under $30^{\circ}C$, relative humidity 95 % condition was conservatively good enough for testing performance of active carbon insitu. The licence change of Yonggwang units 5 and 6 has been completed based on this study.

• The KIPS Transactions:PartA
• /
• v.13A no.4 s.101
• /
• pp.343-350
• /
• 2006

In this paper, wireless internet proxy server clusters are used for the wireless internet because their caching, distillation, and clustering functions are helpful to overcome the limitations and needs of the wireless internet. A wireless Internet proxy server cluster needs a systematic scalability, simple communication structure, cooperative caching, and serving Hot Spot requests. In our former research, we proposed the CD-A structure which can be scalable in a systematic way and has a simple communication structure but it has no cooperative caching. A hash based load balancing can be used to solve the problem, but it can not deal with Hot Spot request problem. In this paper, we proposed a shared storage based wireless internet proxy server cluster which has a systematic scalability, simple communication structure, cooperative caching, and serving Hot Spot requests. The proposed method shares one cache directory and it has advantages: advantages of the existing CD-A structure, cooperative caching, and serving Hot Spot requests. We performed experiments using 16 PCs and experimental results show high performance improvement of the proposed system compared to the existing systems in Hot Spot requests.

• Korean Journal of Air-Conditioning and Refrigeration Engineering
• /
• v.27 no.7
• /
• pp.380-394
• /
• 2015

This article reviews the papers published in the Korean Journal of Air-Conditioning and Refrigeration Engineering during 2014. It is intended to understand the status of current research in the areas of heating, cooling, ventilation, sanitation, and indoor environments of buildings and plant facilities. Conclusions are as follows. (1) The research works on the thermal and fluid engineering have been reviewed as groups of heat and mass transfer, cooling and heating, and air-conditioning, the flow inside building rooms, and smoke control on fire. Research issues dealing with duct and pipe were reduced, but flows inside building rooms, and smoke controls were newly added in thermal and fluid engineering research area. (2) Research works on heat transfer area have been reviewed in the categories of heat transfer characteristics, pool boiling and condensing heat transfer and industrial heat exchangers. Researches on heat transfer characteristics included the results for thermal contact resistance measurement of metal interface, a fan coil with an oval-type heat exchanger, fouling characteristics of plate heat exchangers, effect of rib pitch in a two wall divergent channel, semi-empirical analysis in vertical mesoscale tubes, an integrated drying machine, microscale surface wrinkles, brazed plate heat exchangers, numerical analysis in printed circuit heat exchanger. In the area of pool boiling and condensing, non-uniform air flow, PCM applied thermal storage wall system, a new wavy cylindrical shape capsule, and HFC32/HFC152a mixtures on enhanced tubes, were actively studied. In the area of industrial heat exchangers, researches on solar water storage tank, effective design on the inserting part of refrigerator door gasket, impact of different boundary conditions in generating g-function, various construction of SCW type ground heat exchanger and a heat pump for closed cooling water heat recovery were performed. (3) In the field of refrigeration, various studies were carried out in the categories of refrigeration cycle, alternative refrigeration and modelling and controls including energy recoveries from industrial boilers and vehicles, improvement of dehumidification systems, novel defrost systems, fault diagnosis and optimum controls for heat pump systems. It is particularly notable that a substantial number of studies were dedicated for the development of air-conditioning and power recovery systems for electric vehicles in this year. (4) In building mechanical system research fields, seventeen studies were reported for achieving effective design of the mechanical systems, and also for maximizing the energy efficiency of buildings. The topics of the studies included energy performance, HVAC system, ventilation, and renewable energies, piping in the buildings. Proposed designs, performance performance tests using numerical methods and experiments provide useful information and key data which can improve the energy efficiency of the buildings. (5) The field of architectural environment was mostly focused on indoor environment and building energy. The main researches of indoor environment were related to the evaluation of work noise in tunnel construction and the simulation and development of a light-shelf system. The subjects of building energy were worked on the energy saving of office building applied with window blind and phase change material(PCM), a method of existing building energy simulation using energy audit data, the estimation of thermal consumption unit of apartment building and its case studies, dynamic window performance, a writing method of energy consumption report and energy estimation of apartment building using district heating system. The remained studies were related to the improvement of architectural engineering education system for plant engineering industry, estimating cooling and heating degree days for variable base temperature, a prediction method of underground temperature, the comfort control algorithm of car air conditioner, the smoke control performance evaluation of high-rise building, evaluation of thermal energy systems of bio safety laboratory and a development of measuring device of solar heat gain coefficient of fenestration system.