• 제목/요약/키워드: matrix learning

검색결과 354건 처리시간 0.025초

Universal learning network-based fuzzy control

  • Hirasawa, K.;Wu, R.;Ohbayashi, M.
    • 제어로봇시스템학회:학술대회논문집
    • /
    • 제어로봇시스템학회 1995년도 Proceedings of the Korea Automation Control Conference, 10th (KACC); Seoul, Korea; 23-25 Oct. 1995
    • /
    • pp.436-439
    • /
    • 1995
  • In this paper we present a method to construct fuzzy model with multi-dimension input membership function, which can construct fuzzy inference system on one node of the network directly. This method comes from a common framework called Universal Learning Network (ULN). The fuzzy model under the framework of ULN is called Universal Learning Network-based Fuzzy Inference System (ULNFIS), which possesses certain advantages over other networks such as neural network. We also introduce how to imitate a real system with ULN and a control scheme using ULNFIS.

  • PDF

User Bias Drift Social Recommendation Algorithm based on Metric Learning

  • Zhao, Jianli;Li, Tingting;Yang, Shangcheng;Li, Hao;Chai, Baobao
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제16권12호
    • /
    • pp.3798-3814
    • /
    • 2022
  • Social recommendation algorithm can alleviate data sparsity and cold start problems in recommendation system by integrated social information. Among them, matrix-based decomposition algorithms are the most widely used and studied. Such algorithms use dot product operations to calculate the similarity between users and items, which ignores user's potential preferences, reduces algorithms' recommendation accuracy. This deficiency can be avoided by a metric learning-based social recommendation algorithm, which learns the distance between user embedding vectors and item embedding vectors instead of vector dot-product operations. However, previous works provide no theoretical explanation for its plausibility. Moreover, most works focus on the indirect impact of social friends on user's preferences, ignoring the direct impact on user's rating preferences, which is the influence of user rating preferences. To solve these problems, this study proposes a user bias drift social recommendation algorithm based on metric learning (BDML). The main work of this paper is as follows: (1) the process of introducing metric learning in the social recommendation scenario is introduced in the form of equations, and explained the reason why metric learning can replace the click operation; (2) a new user bias is constructed to simultaneously model the impact of social relationships on user's ratings preferences and user's preferences; Experimental results on two datasets show that the BDML algorithm proposed in this study has better recommendation accuracy compared with other comparison algorithms, and will be able to guarantee the recommendation effect in a more sparse dataset.

딥러닝 기반 항생제 내성균 감염 예측 (Antibiotics-Resistant Bacteria Infection Prediction Based on Deep Learning)

  • 오성우;이한길;신지연;이정훈
    • 한국전자거래학회지
    • /
    • 제24권1호
    • /
    • pp.105-120
    • /
    • 2019
  • 세계보건기구(WHO)를 비롯해 세계 각국의 정부기관은 항생제 오남용에 따른 항생제 내성균 감염에 대해 심각하게 경고하며 이를 예방하기 위한 관리와 감시를 강화하고 있다. 하지만 감염을 확인하기 위한 감염균 배양에 수일의 시간이 소요되면서 격리와 접촉주의를 통한 감염확산 방지 효과가 떨어져 선제적 조치를 위한 신속하고 정확한 예측 및 추정방법이 요구되고 있다. 본 연구는 Electronic Health Records에 포함된 질병 진단내역과 항생제 처방내역을 neural embedding model과 matrix factorization을 통해 embedding 하였고, 이를 활용한 딥러닝 기반분류 예측 모형을 제안하였다. 항생제 내성균 감염의 주요 원인인 질병과 항생제 정보를 embedding하여 환자의 기본정보와 병원이용 정보에 추가했을 때 딥러닝 예측 모형의 f1-score는 0.525에서 0.617로 상승하였고, 딥러닝 모형은 Super Learner와 같은 기존 기계학습 모형보다 더 나은 성능을 보여주었다. 항생제 내성균 감염환자의 특성을 분석한 결과, 감염환자는 동일한 질병을 진단받은 비감염환자에 비교해 J01 계열 항생제 사용이 많았고 WHO 권고기준(DDD)을 크게 벗어나는 오남용 청구사례가 6.3배 이상 높게 나타났으며 항생제 오남용과 항생제 내성균 감염간의 높은 연관성이 발견되었다.

Takagi-Sugeno Fuzzy Model-based Iterative Learning Control Systems: A Two-dimensional System Theory Approach

  • Chu, Jun-Uk;Lee, Yun-Jung
    • 제어로봇시스템학회:학술대회논문집
    • /
    • 제어로봇시스템학회 2001년도 ICCAS
    • /
    • pp.169.3-169
    • /
    • 2001
  • This paper introduces a new approach to analysis of error convergence for a class of iterative learning control systems. First, a nonlinear plant is represented using a Takagi-Sugeno(T-S) fuzzy model. Then each iterative learning controller is designed for each linear plant in the T-S fuzzy model. From the view point of two-dimensional(2-D) system theory, we transform the proposed learning systems to a 2-D error equation, which is also established in the form of T-S fuzzy model. We analysis the error convergence in the sense of induced 2 L -norm, where the effects of disturbances and initial conditions on 2-D error are considered. The iterative learning controller design problem to guarantee the error convergence can be reduced to linear matrix inequality problems. In comparison with others, our learning algorithm ...

  • PDF

비디오 얼굴 식별 성능개선을 위한 다중 심층합성곱신경망 결합 구조 개발 (Development of Combined Architecture of Multiple Deep Convolutional Neural Networks for Improving Video Face Identification)

  • 김경태;최재영
    • 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.

BERT 기반 감성분석을 이용한 추천시스템 (Recommender system using BERT sentiment analysis)

  • 박호연;김경재
    • 지능정보연구
    • /
    • 제27권2호
    • /
    • pp.1-15
    • /
    • 2021
  • 추천시스템은 사용자의 기호를 파악하여 물품 구매 결정을 도와주는 역할을 할 뿐만 아니라, 비즈니스 전략의 관점에서도 중요한 역할을 하기에 많은 기업과 기관에서 관심을 갖고 있다. 최근에는 다양한 추천시스템 연구 중에서도 NLP와 딥러닝 등을 결합한 하이브리드 추천시스템 연구가 증가하고 있다. NLP를 이용한 감성분석은 사용자 리뷰 데이터가 증가함에 따라 2000년대 중반부터 활용되기 시작하였지만, 기계학습 기반 텍스트 분류를 통해서는 텍스트의 특성을 완전히 고려하기 어렵기 때문에 리뷰의 정보를 식별하기 어려운 단점을 갖고 있다. 본 연구에서는 기계학습의 단점을 보완하기 위하여 BERT 기반 감성분석을 활용한 추천시스템을 제안하고자 한다. 비교 모형은 Naive-CF(collaborative filtering), SVD(singular value decomposition)-CF, MF(matrix factorization)-CF, BPR-MF(Bayesian personalized ranking matrix factorization)-CF, LSTM, CNN-LSTM, GRU(Gated Recurrent Units)를 기반으로 하는 추천 모형이며, 실제 데이터에 대한 분석 결과, BERT를 기반으로 하는 추천시스템의 성과가 가장 우수한 것으로 나타났다.

불확실한 로봇 시스템을 위한 적응 반복 학습 제어 및 식별 (An Adaptive Iterative Learning Control and Identification for Uncertain Robotic Systems)

  • 최준영
    • 제어로봇시스템학회논문지
    • /
    • 제10권5호
    • /
    • pp.395-401
    • /
    • 2004
  • We present an AILC(Adaptive Iterative Learning Control) scheme and a sufficient condition for system parameter identification for uncertain robotic systems that perform the same tasks repetitively. It is guaranteed that the joint velocity and position asymptotically converge to the reference joint velocity and position, respectively. In addition, it is proved that a sufficient condition for parameter identification is the PE(Persistent Excitation) condition on the regressor matrix evaluated at the reference trajectory during the operation period. Since the regressor matrix on the reference trajectory can be easily computed prior to the real robot operation, the proposed algorithm provides a useful method to verify whether the parameter error converges to zero or not.

Smart modified repetitive-control design for nonlinear structure with tuned mass damper

  • ZY Chen;Ruei-Yuan Wang;Yahui Meng;Timothy Chen
    • Steel and Composite Structures
    • /
    • 제46권1호
    • /
    • pp.107-114
    • /
    • 2023
  • A new intelligent adaptive control scheme was proposed that combines observer disturbance-based adaptive control and fuzzy adaptive control for a composite structure with a mass-adjustable damper. The most important advantage is that the control structures do not need to know the uncertainty limits and the interference effect is eliminated. Three adjustable parameters in LMI are used to control the gain of the 2D fuzzy control. Binary performance indices with weighted matrices are constructed to separately evaluate validation and training performance using the revalidation learning function. Determining the appropriate weight matrix balances control and learning efficiency and prevents large gains in control. It is proved that the stability of the control system can be ensured by a linear matrix theory of equality based on Lyapunov's theory. Simulation results show that the multilevel simulation approach combines accuracy with high computational efficiency. The M-TMD system, by slightly reducing critical joint load amplitudes, can significantly improve the overall response of an uncontrolled structure.

데이터마이닝 기법들을 통한 제주 안개 예측 방안 연구 (A Study on Fog Forecasting Method through Data Mining Techniques in Jeju)

  • 이영미;배주현;박다빈
    • 한국환경과학회지
    • /
    • 제25권4호
    • /
    • pp.603-613
    • /
    • 2016
  • Fog may have a significant impact on road conditions. In an attempt to improve fog predictability in Jeju, we conducted machine learning with various data mining techniques such as tree models, conditional inference tree, random forest, multinomial logistic regression, neural network and support vector machine. To validate machine learning models, the results from the simulation was compared with the fog data observed over Jeju(184 ASOS site) and Gosan(185 ASOS site). Predictive rates proposed by six data mining methods are all above 92% at two regions. Additionally, we validated the performance of machine learning models with WRF (weather research and forecasting) model meteorological outputs. We found that it is still not good enough for operational fog forecast. According to the model assesment by metrics from confusion matrix, it can be seen that the fog prediction using neural network is the most effective method.

MFMAP: Learning to Maximize MAP with Matrix Factorization for Implicit Feedback in Recommender System

  • Zhao, Jianli;Fu, Zhengbin;Sun, Qiuxia;Fang, Sheng;Wu, Wenmin;Zhang, Yang;Wang, Wei
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제13권5호
    • /
    • pp.2381-2399
    • /
    • 2019
  • Traditional recommendation algorithms on Collaborative Filtering (CF) mainly focus on the rating prediction with explicit ratings, and cannot be applied to the top-N recommendation with implicit feedbacks. To tackle this problem, we propose a new collaborative filtering approach namely Maximize MAP with Matrix Factorization (MFMAP). In addition, in order to solve the problem of non-smoothing loss function in learning to rank (LTR) algorithm based on pairwise, we also propose a smooth MAP measure which can be easily implemented by standard optimization approaches. We perform experiments on three different datasets, and the experimental results show that the performance of MFMAP is significantly better than other recommendation approaches.