• Bulletin of the Korean Chemical Society
• /
• v.33 no.3
• /
• pp.855-861
• /
• 2012

A pioneering version of an on-line management system for high-energy molecules (MS-HEMs) was developed by the ADD and BMDRC in Korea. The current system can manage the physicochemical and explosive properties of virtual and existing HEMs. The on-line MS-HEMs consist of three main routines: management, calculation, and search. The management routine contains a user-friendly interface to store and manage molecular structures and other properties of the new HEMs. The calculation routine automatically calculates a number of compositional and topological molecular descriptors when a new HEM is stored in the MS-HEMs. Physical properties, such as the heat of formation and density, can also be calculated using group additivity methods. In addition, the calculation routine for the impact sensitivity can be used to obtain the safety nature of new HEMs. The impact sensitivity was estimated in a knowledge-based manner using in-house neural network code. The search routine enables general users to find an exact HEM and its properties by sketching a 2D chemical structure, or to retrieve HEMs and their properties by giving a range of properties. These on-line MS-HEMs are expected be powerful tool for deriving novel promising HEMs.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2004.04a
• /
• pp.423-430
• /
• 2004

A new active neuro-control strategy for seismic response reduction using modal states is proposed. In order to apply the neuro-control strategy to the given structural system it is needed to select state variables used as inputs into the neural network. If the degrees of freedom of the analytical model is large, there are so many possible combinations of the state variables. And selecting state variables is very complicated and troublesome task for the designer. In order to avoid this problem, the proposed control system adopts modal states as inputs. Since the modal states contain the information of the whole structural system's behavior, it is proper to use modal states as inputs of the neuro-controller. The simulation results show that the proposed the proposed active neuro-control strategy is quite effective to reduce seismic responses. In addition, the consuming time for training proposed neuro-controller is quite shorter than that for the conventional neuro- controller. The results of this investigation, therefore, indicate that the proposed active neuro-control strategy using modal states as the inputs could be effectively used for control seismically excited structures.

• Journal of Digital Convergence
• /
• v.12 no.1
• /
• pp.319-324
• /
• 2014

In recent years, high-performance flying drones attract attention for many peoples. In particular, the drone equipped with multi-rotor is expanding its range of utilization in video imaging, aerial rescue, logistics, monitoring, measurement, military field, etc. However, the control function of its controller is very simple. In this study, using a G-sensor mounted on a mobile device, implements an enhanced controller to control flying drones through the intuitive gesture of user. The implemented controller improves the gesture recognition performance using a neural network algorithm.

• IEIE Transactions on Smart Processing and Computing
• /
• v.2 no.6
• /
• pp.323-331
• /
• 2013

An extreme learning machine (ELM) is an efficient learning algorithm that is based on the generalized single, hidden-layer feed-forward networks (SLFNs), which perform well in classification applications. Many studies have demonstrated its superiority over the existing classical algorithms: support vector machine (SVM) and BP neural network. This paper presents a novel face recognition approach based on a multi-class project extreme learning machine (MPELM) classifier and 2D Gabor transform. First, all face image features were extracted using 2D Gabor filters, and the MPELM classifier was used to determine the final face classification. Two well-known face databases (CMU-PIE and ORL) were used to evaluate the performance. The experimental results showed that the MPELM-based method outperformed the ELM-based method as well as other methods.

• Proceedings of The KACE
• /
• 2017.08a
• /
• pp.201-204
• /
• 2017

빅 데이터의 시대가 도래하면서 이전보다 데이터로부터 유의미한 정보를 추출하는 것에 대한 연구가 활발하게 진행되고 있다. 딥러닝은 텍스트, 이미지, 동영상 등 다양한 데이터에 대한 학습을 가능하게 할 뿐만 아니라 높은 학습 정확도를 보임으로써 차세대 머선러닝 기술로 각광 받고 있다. 그러나 딥러닝은 일반적으로 학습해야하는 데이터가 많을 뿐만 아니라 학습에 요구되는 시간이 매우 길다. 또한 데이터의 전처리 수준과 학습 모델 튜닝에 의해 학습정확도가 크게 영향을 받기 때문에 활용이 어렵다. 딥러닝에서 학습에 요구되는 데이터의 양과 연산량이 많아지면서 분산 처리 프레임워크 기반 분산 학습을 통해 학습 정확도는 유지하면서 학습시간을 단축시키는 사례가 많아지고 있다. 본 연구에서는 범용 분산 처리 프레임워크인 아파치 스파크에서 데이터 병렬화 기반 분산 학습 모델을 활용하여 모바일 빅 데이터 분석을 위한 딥러닝을 구현한다. 딥러닝을 구현할 때 분산학습을 통해 학습 속도를 높이면서도 학습 정확도를 높이기 위한 모델 튜닝 방법을 연구한다. 또한 스파크의 분산 병렬처리 효율을 최대한 끌어올리기 위해 파티션 병렬 최적화 기법을 적용하여 딥러닝의 학습속도를 향상시킨다.

• Journal of Industrial Convergence
• /
• v.18 no.3
• /
• pp.53-61
• /
• 2020

Travel speed is an important parameter for measuring road traffic and incident detection system. In this paper I suggests a model developed for estimating reliable and accurate average roadway link travel speeds using image processing sensor. This method extracts the vehicles from the video image from CCTV, tracks the moving vehicles using deep neural network, and extracts traffic information such as link travel speeds and volume. The algorithm estimates link travel speeds using a robust data-fusion procedure to provide accurate link travel speeds and traffic information to the public. In the field tests, the new model performed better than existing methods.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.3
• /
• pp.51-55
• /
• 2020

In this paper, an improved FAM is proposed by adopting similarity learning in the existing FAM (Fuzzy Associative Memory) used in image restoration. Image restoration refers to the recovery of the latent clean image from its noise-corrupted version. In serious application like face recognition, this process should be noise-tolerant, robust, fast, and scalable. The existing FAM is a simple single layered neural network that can be applied to this domain with its robust fuzzy control but has low capacity problem in real world applications. That similarity measure is implied to the connection strength of the FAM structure to minimize the root mean square error between the recovered and the original image. The efficacy of the proposed algorithm is verified with significant low error magnitude from random noise in our experiment.

• Journal of KIISE
• /
• v.44 no.5
• /
• pp.503-509
• /
• 2017

In this paper, the copy mechanism and input feeding are applied to recurrent neural network(RNN)-search model in a Korean-document summarization in an end-to-end manner. In addition, the performances of the document summarizations are compared according to the model and the tokenization format; accordingly, the syllable-unit, morpheme-unit, and hybrid-unit tokenization formats are compared. For the experiments, Internet newspaper articles were collected to construct a Korean-document summary data set (train set: 30291 documents; development set: 3786 documents; test set: 3705 documents). When the format was tokenized as the morpheme-unit, the models with the input feeding and the copy mechanism showed the highest performances of ROUGE-1 35.92, ROUGE-2 15.37, and ROUGE-L 29.45.

• International Journal of Naval Architecture and Ocean Engineering
• /
• v.9 no.4
• /
• pp.373-381
• /
• 2017

A Collaborative Optimization (CO) methodology for ring-stiffened composite material pressure hull of underwater vehicle is proposed. Structural stability and material strength are both examined. Lamination parameters of laminated plates are introduced to improve the optimization efficiency. Approximation models are established based on the Ellipsoidal Basis Function (EBF) neural network to replace the finite element analysis in layout optimizers. On the basis of a two-level optimization, the simultaneous structure material collaborative optimization for the pressure vessel is implemented. The optimal configuration of metal liner and frames and composite material is obtained with the comprehensive consideration of structure and material performances. The weight of the composite pressure hull decreases by 30.3% after optimization and the validation is carried out. Collaborative optimization based on the lamination parameters can optimize the composite pressure hull effectively, as well as provide a solution for low efficiency and non-convergence of direct optimization with design variables.

• Annual Conference on Human and Language Technology
• /
• 2017.10a
• /
• pp.137-142
• /
• 2017

원격 지도 학습은 자동으로 매우 큰 코퍼스와 지식베이스 간의 주석 데이터를 생성하여 기계 학습에 필요한 학습 데이터를 사람의 손을 빌리지 않고 저렴한 비용으로 만들 수 있어, 많은 연구들이 관계 추출 문제를 해결하기 위해 원격 지도 학습 방법을 적용하고 있다. 그러나 기존 연구들에서는 모델 학습의 입력으로 사용되는 단어 임베딩에서 단어의 동형이의어 성질을 반영하지 못한다는 단점이 있다. 때문에 서로 다른 의미를 가진 동형이의어가 하나의 임베딩 값을 가지다 보니, 단어의 의미를 정확히 파악하지 못한채 관계 추출 모델을 학습한다고 볼 수 있다. 본 논문에서는 원격 지도 학습 기반 관계 추출 모델에 다중-어의 단어 임베딩을 적용한 모델을 제안한다. 다중-어의 단어 임베딩 학습을 위해 어의 중의성 해소 모듈을 활용하였으며, 관계 추출 모델은 문장 내 주요 특징을 효율적으로 파악하는 모델인 CNN과 PCNN을 활용하였다. 본 논문에서 제안하는 다중-어의 단어 임베딩 적용 관계추출 모델의 성능을 평가하기 위해 추가적으로 2가지 방식의 단어 임베딩을 학습하여 비교 평가를 수행하였고, 그 결과 어의 중의성 해소 모듈을 활용한 단어 임베딩을 활용하였을 때 관계추출 모델의 성능이 향상된 결과를 보였다.