• Geomechanics and Engineering
• /
• 제13권5호
• /
• pp.845-860
• /
• 2017

It is very important to study displacement prediction in geotechnical engineering. Nowadays, the grey system method, time series analysis method and artificial neural network method are three main methods. Based on the brief introduction, the three methods are analyzed comprehensively. Their merits and demerits, applied ranges are revealed. To solve the shortcomings of the artificial neural network method, a new prediction method based on new evolutionary neural network is proposed. Finally, through two real engineering applications, the analysis of three main methods and the new evolutionary neural network method all have been verified. The results show that, the grey system method is a kind of exponential approximation to displacement sequence, and time series analysis is linear autoregression approximation, while artificial neural network is nonlinear autoregression approximation. Thus, the grey system method can suitably analyze the sequence, which has the exponential law, the time series method can suitably analyze the random sequence and the neural network method almostly can be applied in any sequences. Moreover, the prediction results of new evolutionary neural network method is the best, and its approximation sequence and the generalization prediction sequence are all coincided with the real displacement sequence well. Thus, the new evolutionary neural network method is an acceptable method to predict the measurement displacements of geotechnical engineering.

• 대한기계학회논문집
• /
• 제18권7호
• /
• pp.1881-1897
• /
• 1994

This paper presents a neural network-based computational scheme to generate the optimized robotic assembly sequence for an assembly product consisting of a number of parts. An assembly sequence is considered to be optimal when it meets a number of conditions : it must satisfy assembly constraints, keep the stability of in-process subassemblies, and minimize assembly cost. To derive such an optimal sequence, we propose a scheme using both the Hopfield neural network and the expert system. Based upon the inferred precedence constraints and the assembly costs from the expert system, we derive the evolution equation of the network. To illustrate the suitability of the proposed scheme, a case study is presented for industrial product of an electrical relay. The result is compared with that obtained from the expert system.

• 스마트미디어저널
• /
• 제7권4호
• /
• pp.17-23
• /
• 2018

자동 띄어쓰기 특성을 효과적으로 처리할 수 있는 LSTM(Long Short-Term Memory Neural Networks) 기반의 RNN 모델을 제시하고 적용한 결과를 분석하였다. 문장이 길거나 일부 노이즈가 포함된 경우에 신경망 학습이 쉽지 않은 문제를 해결하기 위하여 입력 데이터 형식과 디코딩 데이터 형식을 정의하고, 신경망 학습에서 드롭아웃, 양방향 다층 LSTM 셀, 계층 정규화 기법, 주목 기법(attention mechanism)을 적용하여 성능을 향상시키는 방법을 제안하였다. 학습 데이터로는 세종 말뭉치 자료를 사용하였으며, 학습 데이터가 부분적으로 불완전한 띄어쓰기가 포함되어 있었음에도 불구하고, 대량의 학습 데이터를 통해 한글 띄어쓰기에 대한 패턴이 의미 있게 학습되었다. 이것은 신경망에서 드롭아웃 기법을 통해 학습 모델의 오버피팅이 되지 않도록 함으로써 노이즈에 강한 모델을 만들었기 때문이다. 실험결과로 LSTM sequence-to-sequence 모델이 재현율과 정확도를 함께 고려한 평가 점수인 F1 값이 0.94로 규칙 기반 방식과 딥러닝 GRU-CRF보다 더 높은 성능을 보였다.

• 정보과학회 논문지
• /
• 제45권2호
• /
• pp.134-140
• /
• 2018

채팅 시스템은 사람의 말을 기계가 이해하고 적절한 응답을 하는 시스템이다. 채팅 시스템은 사용자의 간단한 정보 검색 질문에 대답해야 하는 경우가 있다. 그러나 기존의 생성 채팅 시스템들은 질의응답에 필요한 정보인 지식 개체(트리플 형태 지식에서의 주어와 목적어)의 임베딩을 고려하지 않아 발화에 나타나는 지식 개체가 다르더라도 같은 형태의 답변이 생성되었다. 본 논문에서는 생성 기반 채팅 시스템의 질의응답 정확도를 향상시키기 위한 지식 임베딩 방법을 제안한다. 개체와 유의어의 지식 임베딩을 위해 샴 순환 신경망을 사용하며 이를 이용해 주어와 술어를 인코딩 하고 목적어를 디코딩하는 sequence-to-sequence 모델의 성능을 향상 시켰다. 자체 구축한 채팅데이터를 통한 실험에서 제안된 임베딩 방법은 종래의 합성곱 신경망을 통한 임베딩 방법 보다 12.48% 높은 정확도를 보였다.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2003년도 Proceedings of ACRS 2003 ISRS
• /
• pp.429-431
• /
• 2003

The induced learning retardation method involves the temporary inhibition of the artificial neural network’s active units from participating in the error reduction process during training. This stimulates the less active units to contribute significantly to reduce the network error. However, some less active units are not sensitive to stimulation making them almost useless. The network can then be pruned by removing the less active units to make it smaller and more efficient. This study focuses on making the network more efficient and accurate by developing the induced learning retardation and pruning sequence training method. The developed procedure results to faster learning and more accurate artificial neural network for satellite image classification.

• 한국멀티미디어학회논문지
• /
• 제22권6호
• /
• pp.655-664
• /
• 2019

In this paper, we propose a novel way of combining multiple deep convolutional neural network (DCNN) architectures which work well for accurate video face identification by adopting a serial combination of 3D and 2D DCNNs. The proposed method first divides an input video sequence (to be recognized) into a number of sub-video sequences. The resulting sub-video sequences are used as input to the 3D DCNN so as to obtain the class-confidence scores for a given input video sequence by considering both temporal and spatial face feature characteristics of input video sequence. The class-confidence scores obtained from corresponding sub-video sequences is combined by forming our proposed class-confidence matrix. The resulting class-confidence matrix is then used as an input for learning 2D DCNN learning which is serially linked to 3D DCNN. Finally, fine-tuned, serially combined DCNN framework is applied for recognizing the identity present in a given test video sequence. To verify the effectiveness of our proposed method, extensive and comparative experiments have been conducted to evaluate our method on COX face databases with their standard face identification protocols. Experimental results showed that our method can achieve better or comparable identification rate compared to other state-of-the-art video FR methods.

• 인지과학
• /
• 제18권2호
• /
• pp.179-191
• /
• 2007

목적 지향 대화에서 사용자의 의도는 화행과 개념열의 쌍으로 구성된 영역행위로 표현될 수 있다. 사용자 발화에 대한 영역행위 예측은 음성 인식 오류를 보정하는데 유용하며, 시스템 발화에 대한 영역행위 예측은 유연한 응답 생성에 유용하다. 본 논문에서는 신경망을 이용하여 영역행위를 예측하는 모델을 제안한다. 제안 모델은 대화 이력 벡터와 현재 영역행위를 신경망의 입력으로 사용하여 다음 영역행위를 예측한다. 실험 결과, 제안 모델은 화행 예측과 개념열 예측에서 각각 80.02%, 82.09%의 정확률을 보였다.

• Journal of Information Processing Systems
• /
• 제9권2호
• /
• pp.259-270
• /
• 2013

A speaker's intentions can be represented by domain actions (domain-independent speech act and domain-dependent concept sequence pairs). Therefore, it is essential that domain actions be determined when implementing dialogue systems because a dialogue system should determine users' intentions from their utterances and should create counterpart intentions to the users' intentions. In this paper, a neural network model is proposed for classifying a user's domain actions and planning a system's domain actions. An integrated neural network model is proposed for simultaneously determining user and system domain actions using the same framework. The proposed model performed better than previous non-integrated models in an experiment using a goal-oriented dialogue corpus. This result shows that the proposed integration method contributes to improving domain action determination performance.

• 한국음향학회지
• /
• 제39권5호
• /
• pp.476-482
• /
• 2020

입력열과 출력열의 길이가 다른 경우 attention 기법을 이용한 sequence-to-sequence 심층 신경망이 우수한 성능을 보인다. 그러나, 출력열의 길이에 비해서 입력열의 길이가 너무 긴 경우, 그리고 하나의 출력값에 해당하는 입력열의 특성이 변화하는 경우, 하나의 문맥 벡터(context vector)를 사용하는 기존의 attention 방법은 적당하지 않을 수 있다. 본 논문에서는 이러한 문제를 해결하기 위해서 입력열의 왼쪽 부분과 오른쪽 부분을 각각 개별적으로 처리할 수 있는 두 개의 문맥 벡터를 사용하는 이중 attention 기법을 제안한다. 제안한 방법의 효율성은 TIMIT 데이터를 사용한 음성 인식 실험을 통하여 검증하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권2호
• /
• pp.538-561
• /
• 2020

Recommender Systems (RecSys) have a major role in e-commerce for recommending products, which they may like for every user and thus improve their business aspects. Although many types of RecSyss are there in the research field, the state of the art RecSys has focused on finding the user similarity based on sequence (e.g. purchase history, movie-watching history) analyzing and prediction techniques like Recurrent Neural Network in Deep learning. That is RecSys has considered as a sequence prediction problem. However, evaluation of similarities among the customers is challenging while considering temporal aspects, context and multi-component ratings of the item-records in the customer sequences. For addressing this issue, we are proposing a Deep Learning based model which learns customer similarity directly from the sequence to sequence similarity as well as item to item similarity by considering all features of the item, contexts, and rating components using Dynamic Temporal Warping(DTW) distance measure for dynamic temporal matching and 2D-GRU (Two Dimensional-Gated Recurrent Unit) architecture. This will overcome the limitation of non-linearity in the time dimension while measuring the similarity, and the find patterns more accurately and speedily from temporal and spatial contexts. Experiment on the real world movie data set LDOS-CoMoDa demonstrates the efficacy and promising utility of the proposed personalized RecSys architecture.