• Journal of Information Processing Systems
• /
• 제19권5호
• /
• pp.688-701
• /
• 2023

The conventional methods of network intrusion detection system (NIDS) cannot measure the trend of intrusiondetection targets effectively, which lead to low detection accuracy. In this study, a NIDS method which based on a deep neural network in a big-data environment is proposed. Firstly, the entire framework of the NIDS model is constructed in two stages. Feature reduction and anomaly probability output are used at the core of the two stages. Subsequently, a convolutional neural network, which encompasses a down sampling layer and a characteristic extractor consist of a convolution layer, the correlation of inputs is realized by introducing bidirectional long short-term memory. Finally, after the convolution layer, a pooling layer is added to sample the required features according to different sampling rules, which promotes the overall performance of the NIDS model. The proposed NIDS method and three other methods are compared, and it is broken down under the conditions of the two databases through simulation experiments. The results demonstrate that the proposed model is superior to the other three methods of NIDS in two databases, in terms of precision, accuracy, F1- score, and recall, which are 91.64%, 93.35%, 92.25%, and 91.87%, respectively. The proposed algorithm is significant for improving the accuracy of NIDS.

• International Journal of Advanced Culture Technology
• /
• 제11권4호
• /
• pp.393-405
• /
• 2023

Accurate prediction of wind speed and power is vital for enhancing the efficiency of wind energy systems. Numerous solutions have been implemented to date, demonstrating their potential to improve forecasting. Among these, deep learning is perceived as a revolutionary approach in the field. However, despite their effectiveness, the noise present in the collected data remains a significant challenge. This noise has the potential to diminish the performance of these algorithms, leading to inaccurate predictions. In response to this, this study explores a novel feature engineering approach. This approach involves altering the data input shape in both Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) and Autoregressive models for various forecasting horizons. The results reveal substantial enhancements in model resilience against noise resulting from step increases in data. The approach could achieve an impressive 83% accuracy in predicting unseen data up to the 24th steps. Furthermore, this method consistently provides high accuracy for short, mid, and long-term forecasts, outperforming the performance of individual models. These findings pave the way for further research on noise reduction strategies at different forecasting horizons through shape-wise feature engineering.

• 방송공학회논문지
• /
• 제25권2호
• /
• pp.200-207
• /
• 2020

CNN 기반 인공신경망은 영상 분류, 객체 인식, 화질 개선 등 다양한 분야에서 뛰어난 성능을 보이고 있다. 그러나, 많은 응용에서 딥러닝(Deep Learning) 모델의 복잡도 및 연산량이 방대해짐에 따라 IoT 기기 및 모바일 환경에 적용하기에는 제한이 따른다. 따라서 기존 딥러닝 모델의 성능을 유지하면서 모델 크기를 줄이는 인공신경망 압축 기법이 연구되고 있다. 본 논문에서는 인공신경망 압축기법을 통하여 원본 CNN 모델을 압축하고, 압축된 모델을 임베디드 시스템 환경에서 그 성능을 검증한다. 성능 검증을 위해 인공지능 지원 맞춤형 칩인 QCS605를 내장한 임베디드 보드에서 카메라로 입력한 영상에 대해서 원 CNN 모델과 압축 CNN 모델의 분류성능과 추론시간을 비교 분석한다. 본 논문에서는 이미지 분류 CNN 모델인 MobileNetV2, ResNet50 및 VGG-16에 가지치기(pruning) 및 행렬분해의 인공신경망 압축 기법을 적용하였고, 실험결과에서 압축된 모델이 원본 모델 분류 성능 대비 2% 미만의 손실에서 모델의 크기를 1.3 ~ 11.2배로 압축했을 뿐만 아니라 보드에서 추론시간과 메모리 소모량을 각각 1.2 ~ 2.1배, 1.2 ~ 3.8배 감소함을 확인했다.

• 한국항행학회논문지
• /
• 제23권1호
• /
• pp.84-89
• /
• 2019

본 연구는 복합재 항공기의 비행 데이터를 활용한 기계학습 기반 구조건전성 모니터링 시스템 연구의 예비 연구이다. 본 연구에서는 구조건전성 모니터링에 이용되기에 가장 적합한 기계학습 알고리즘을 선별하고, 실 기체 데이터에 대한 적용을 위해 차원 축소를 수행하였다. 이를 위해 외팔보를 통해 모사된 항공기 날개 구조와 부가 질량을 통해 손상 모사 실험을 진행하고, 분류 알고리즘을 통해 데이터를 손상의 위치와 정도에 따라 구분하였다. 이를 위해 FBG (fiber bragg grating) 센서를 부착한 외팔보의 진동 실험을 통해 정상상태와 12개의 손상상태에 대한 데이터를 취득하고, MATLAB 환경에서 tree, discriminant, SVM (support vector machine), kNN, ensemble 알고리즘의 비교와 파라미터 튜닝을 통해 가장 적합한 알고리즘을 도출하였다. 또한 NCA (neighborhood component analysis)를 이용한 특징 선택을 통해, 실 기체에서 나올 수 있는 고차원 데이터의 관리를 위해 필요한 차원 축소를 수행하였다. 그 결과, quadratic SVM이 NCA를 적용하지 않은 모델에서 98.7%, NCA를 적용한 모델에서 95.9%로 가장 높은 정답률을 보였다. 또한 NCA 적용 후 모델의 예측 속도, 학습 시간, 용량이 모두 향상되었다.

• 동의신경정신과학회지
• /
• 제10권1호
• /
• pp.17-38
• /
• 1999

The effects of Sahyangsohapwon on the affective reactivity of rats were studied with open-field behavior. Sample group was treated with the medicine for 8 weeks, whereas control group was treated with the vehicle. The effects of Sahyangsohapwon on the enhancement of learning and memory of AD model rats were studied with two-way avoidance task. Sample group electrically lesioned on nbM(nucleus basalis of Meynert) was treated with the medicine for 8 weeks, whereas control group with nbM lesion and sham group with the sham operation were treated with the vehicle. 1. In the open-field behavior task, the start latency from start box was measured $27.08{\pm}7.51sec$ in control group, $23.15{\pm}5.98sec$ in sample group. Rats in sample group showed a tendency of shortened latency going out to a strange place compared with those in control group, but with no statistical significance(p>0.05). 2. In the open-field behavior task, the number of locomotion crossing the grid lines was measured $84.54{\pm}3.55$ in control group, $116.93{\pm}6.41$ in sample group. There was an increased locomotion in sample group compared with control group with statistical significance(p<0.01). This can be interpreted as rats in sample group showed lowerd anxiety under a strange environment. 3. In the open-field behavior task, the rearing number was measured $7.46{\pm}0.57$ in control group, $10.13{\pm}0.95$ in sample group. There was an increased rearing in sample group compared with control group with statistical significance(p<0.05). This can also be interpreted as rats in sample group showed lowerd anxiety under a strange environment. 4. In the open-field behavior task, the number of crossing behavior was measured $5.54{\pm}1.50$ in control group, $9.20{\pm}1.67$ in sample group. There was a increasing tendency of crossing behavior in sample group compared with control group, but with no statistical significance(p<0.05). 5. In the open-field behavior task, the total activity was measured $97.54{\pm}4.70$ in control group, $136.27{\pm}792$ in sample group. There was an increased total activity in sample group compared with control group with statistical significance(p<0.01). This can also be interpreted as rats in sample group showed lowerd anxiety under a strange environment. 6. In the analysis of effects on the learning and memory in AD model rats with two-way avoidance task, the response latency was measured $6717{\pm}134msec$ in the 1st session, $5416{\pm}160msec$ in the 2nd session, $5252{\pm}148msec$ in the 3rd session in control group. It was measured $6724{\pm}155msec$ in the 1st session, $4642{\pm}139msec$ in the 2nd session, $4914{\pm}148msec$ in the 3rd session in sample group and $4357{\pm}144msec$ in the 1st session, $3125{\pm}115msec$ in the 2nd session, $3091{\pm}98msec$ in the 3rd session in sham group. There were differences between sham group and nbM lesioned groups with statistical significance in post hoc analysis(p<0.000). And in the 2nd session, there was a reduction of latency in sample group compared with control group with statistical significance (p<0.000). This showed that sample group had better learning capacity than control group. 7. In the analysis of effects on the learning and memory in AD model rats with two-way avoidance task, the number of avoidance response was measured $5.85{\pm}1.41$ in the 1st session, $14.23{\pm}2.89$ in the 2nd session, $15.69{\pm}2.56$ in the 3rd session in control group. It was measured $7.92{\pm}1.94$ in the 1st session, $16.83{\pm}2.29$ in the 2nd session, $15.42{\pm}2.81$ in the 3rd session in sample group and $14.38{\pm}1.62$ in the 1st session, $22.88{\pm}0.89$ in the 2nd session, $23.88{\pm}1.64$ in the 3rd session in sham group. There were differences between sham group and nbM lesioned groups with statistical significance in post hoc analysis(p<0.001). But between control and sample group, there was no significant difference. With the experimental results above, Sahyangsohapwon can be supposed to have the enhancing effects on the affect reactivity and learning with memory of AD model rats induced by electrolyte injury of nbM.

• Molecules and Cells
• /
• 제38권9호
• /
• pp.796-805
• /
• 2015

Gintonin is a novel ginseng-derived lysophosphatidic acid (LPA) receptor ligand. Oral administration of gintonin ameliorates learning and memory dysfunctions in Alzheimer's disease (AD) animal models. The brain cholinergic system plays a key role in cognitive functions. The brains of AD patients show a reduction in acetylcholine concentration caused by cholinergic system impairments. However, little is known about the role of LPA in the cholinergic system. In this study, we used gintonin to investigate the effect of LPA receptor activation on the cholinergic system in vitro and in vivo using wild-type and AD animal models. Gintonin induced $[Ca^{2+}]_i $ transient in cultured mouse hippocampal neural progenitor cells (NPCs). Gintonin-mediated $[Ca^{2+}]_i $ transients were linked to stimulation of acetylcholine release through LPA receptor activation. Oral administration of gintonin-enriched fraction (25, 50, or 100 mg/kg, 3 weeks) significantly attenuated scopolamine-induced memory impairment. Oral administration of gintonin (25 or 50 mg/kg, 1 2 weeks) also significantly attenuated amyloid-${\beta}$ protein ($A{\beta}$)-induced cholinergic dysfunctions, such as decreased acetylcholine concentration, decreased choline acetyltransferase (ChAT) activity and immunoreactivity, and increased acetylcholine esterase (AChE) activity. In a transgenic AD mouse model, long-term oral administration of gintonin (25 or 50 mg/kg, 3 months) also attenuated AD-related cholinergic impairments. In this study, we showed that activation of G protein-coupled LPA receptors by gintonin is coupled to the regulation of cholinergic functions. Furthermore, this study showed that gintonin could be a novel agent for the restoration of cholinergic system damages due to $A{\beta}$ and could be utilized for AD prevention or therapy.

• 대한전자공학회논문지TC
• /
• 제42권7호
• /
• pp.27-34
• /
• 2005

Iterative detection은 additive white Gaussian noise(AWGN) channel의 경우 interleaver들을 포함한 조합유한상태머신(concatenated Finite State Machine)들에 대해 근사적으로 optimal solution에 가깝다는 것이 입증되었습니다. 수신단에서 정확한 채널 상태 정보(perfect channel state information)가 얻어질 수 없는 경우 adaptive Iterative detection이 시간적으로 변하거나 또는 부정확한 채널 변수를 다루기위해 필요합니다. Iterative detection과 adaptive iterative detection대한 기본 building block은 각각 Soft-Input Soft-Output (SISO)와adaptive SISO (A-SISO)입니다. SISO와 A-SISO의 complexity은 state memory나 channel memory에 비례해서 지수적으로 증가합니다. 본 논문에서는 Reduced State SISO (RS-SISO) 알고리즘이 A-SISO의 complexity 감소를 위해 적용되어 fading ISI channel을 통한 serially concatenated CPM의 성능이 adaptive iterative detection을 이용하면 터보 코드 같은 성능을 나타내는 것과 또한 RS-A-SISO system이 큰 iterative detection gain을 가지는 것을 보였습니다. RS-A-SISO 알고리즘에 대한 다양한 design option들의 성능을 평가하였으며 성능과 complexity를 비교하였습니다. 또한 보통 AWGN 채널에서 사용되어지는 density evolution 분석기법이 주파수 선택적인 페이딩 채널에서 RS-A-SISO 시스템에서도 좋은 분석기법임을 보였습니다

• 대한한의학회지
• /
• 제35권4호
• /
• pp.10-23
• /
• 2014

Objectives: This study evaluated the effects of Guibi-tang (GBT) on neuronal apoptosis and cognitive impairment induced by beta amyloid ($A{\beta}$), (1-42) injection in the hippocampus of ICR mice. Methods: $A{\beta}$ (1-42) was injected unilaterally into the lateral ventricle using a Hamilton syringe and micropump ($2{\mu}g/3{\mu}{\ell}$, $0.6{\mu}{\ell}/min$). Water extract of GBT was administered orally once a day (500 mg/kg) for 3 weeks after the $A{\beta}$ (1-42) injection. Acquisition of learning and retention of memory were tested using the Morris water maze. Neuronal damage and $A{\beta}$ accumulation in the hippocampus was observed using cresyl violet and Congo red staining. The anti-apoptotic effect of GBT was evaluated using TUNEL labeling in the hippocampus. Results: GBT significantly shortened the escape latencies during acquisition training trials. GBT significantly increased the number of target headings to the platform site, the swimming time spent in the target quadrant, and significantly shortened the time for the 1st target heading during the retention test trial. GBT significantly attenuated the reduction in thickness and number of CA1 neurons, and $A{\beta}$ accumulation in the hippocampus produced by $A{\beta}$ (1-42) injection. GBT significantly reduced the number of TUNEL-labeled neurons in the hippocampus. Conclusion: These results suggest that GBT improved cognitive impairment by reducing neuronal apoptosis and $A{\beta}$ accumulation in the hippocampus. GBT may be a beneficial herbal formulation in treating cognitive impairment including Alzheimer's disease.

• 대한전자공학회논문지SD
• /
• 제43권3호
• /
• pp.48-58
• /
• 2006

본 논문은 가산기 기반 DA(Distributed Arithmetic: 분산 산술연산)구조로서 ROM과 같은 일반적인 메모리가 사용되지 않는 8x8의 2차원 DCT(Discrete Cosine Transform)/IDCT(Inverse DCT) 프로세서를 제안 설계하였다. 제안된 논문은 DCT와 IDCT의 계수 행렬에서 하드웨어를 줄이기 위해 계수 행렬의 홀수 부분을 공유하였고, 2차원 DCT/IDCT 프로세서의 계수 연산을 위해 단지 29개의 가산기만을 사용하였다. 이는 8x8 1차원 DCT NEDA(NEw DA)구조에서의 가산기 수 보다 48.6%를 감소 시켰다. 또한, 기존의 전치메모리와는 다른 새로운 전치네트워크 구조를 제안하였다. 제안된 전치네트워크 구조에서는 전치메모리 블록 대신 하드웨어를 줄이기 위해 레지스터 형태의 새로운 레지스터 블록 전치네트워크 형태를 제안하였다. 제안된 전치네트워크 블록은 64개의 레지스터를 사용하며, 이는 일반적인 메모리를 사용하는 기존의 전치메모리 구조에 사용된 트랜지스터 수 보다 18%가 감소하였다. 또한 처리율 향상을 위해 새롭게 적용되고 있는 방식으로, 입력 데이터에 대해 매 클럭 주기마다 8개의 화소데이터를 받아서 8개의 화소데이터를 처리하도록 하여 출력하는 비트 병렬화 구조로 설계하였다.

• 한국정보과학회논문지:컴퓨팅의 실제 및 레터
• /
• 제9권1호
• /
• pp.33-46
• /
• 2003

HLR system은 이동전화 망에서 지속적으로 변하는 개별 가입자의 위치 정보를 관리한다. 이를 수행하기 위해, HLR database system은 table 관리 기능과 색인 관리 기능, 그리고 백업 관리 기능을 제공한다. 본 논문에서는, 이동 전화 번호(MDN : Mobile Directory Number)를 위한 적절한 객인 기법으로서 이단계 색인 기법의 사용과, 단말번호(ESN : Electronic Serial Number)를 위한 버켓 연결 해슁 기법을 제안한다. 이동 전화 번호(MDN)와 단말번호(ESN)는 HLR database system에서 key로 사용된다. 또한 HLR database transaction의 특성을 고려한 효율적인 백업 방법을 제안한다. 이단계 색인 기법은 기존의 T 트리 색인 기법보다 검색 속도와 기억 공간 사용 효율 측면에서 우수하다. 버켓 연결 해슁 기법은 기존의 변형된 선형 해슁 기법보다 삽입과 삭제 시의 오버헤드가 적다. 제안한 백업 방법에서는, 빈번한 위치 등록 기능 수행으로 인해 야기되는 성능 저하 문제론 해결하기 위해 두가지 종류의 갱신 플래그를 사용하였다. 100만 가입자 수용시. 제안 기법을 사용하게 되련 기존 기법보다 메모리 사용량 절감(62% 이상), 디렉토리 증가 작업(25만 번 이상)제거, 백업 작업 감소(80% 이상)를 제공받게 된다.