1,726,277 research outputs found
Forecasting Daily Gas Load with OIHF-Elman Neural Network
No full textAbstractTo improve the forecasting accuracy, a model for forecasting daily gas load with OIHF-Elman network involving factors such as weather, temperature and data type is proposed. Compared with the conventional Elman network, OIHF-Elman network considers not only the hidden level feedback but also the output level feedbacks. Therefore more information from limited sampling spots is collected and utilized. The simulation results show that OIHF-Elman network performs better than Elman network in terms of accuracy given limited sampling points. The new model also improves the generalization of information and can be used to forecast the daily gas load successfully
Elman and SGWO-Elman optimization process.
No full textTraditional neural networks used gradient descent methods to train the network structure, which cannot handle complex optimization problems. We proposed an improved grey wolf optimizer (SGWO) to explore a better network structure. GWO was improved by using circle population initialization, information interaction mechanism and adaptive position update to enhance the search performance of the algorithm. SGWO was applied to optimize Elman network structure, and a new prediction method (SGWO-Elman) was proposed. The convergence of SGWO was analyzed by mathematical theory, and the optimization ability of SGWO and the prediction performance of SGWO-Elman were examined using comparative experiments. The results show: (1) the global convergence probability of SGWO was 1, and its process was a finite homogeneous Markov chain with an absorption state; (2) SGWO not only has better optimization performance when solving complex functions of different dimensions, but also when applied to Elman for parameter optimization, SGWO can significantly optimize the network structure and SGWO-Elman has accurate prediction performance.</div
Elman neural network structure.
No full textTraditional neural networks used gradient descent methods to train the network structure, which cannot handle complex optimization problems. We proposed an improved grey wolf optimizer (SGWO) to explore a better network structure. GWO was improved by using circle population initialization, information interaction mechanism and adaptive position update to enhance the search performance of the algorithm. SGWO was applied to optimize Elman network structure, and a new prediction method (SGWO-Elman) was proposed. The convergence of SGWO was analyzed by mathematical theory, and the optimization ability of SGWO and the prediction performance of SGWO-Elman were examined using comparative experiments. The results show: (1) the global convergence probability of SGWO was 1, and its process was a finite homogeneous Markov chain with an absorption state; (2) SGWO not only has better optimization performance when solving complex functions of different dimensions, but also when applied to Elman for parameter optimization, SGWO can significantly optimize the network structure and SGWO-Elman has accurate prediction performance.</div
Translation : Benjamin A. Elman, A Cultural History of Civil Examinations in Late Imperial China, chapter 6, 2
No full text本訳文は、Benjamin A. Elman. A Cultural History of Civil Examinations in Late Imperial China. Berkeley: University of California Press. （2000）の第六章の翻訳である。departmental bulletin pape
Design of Speed Loop to Image-stabilization Platform Based on Elman Neural Network
No full text稳像平台速度环的性能直接影响成像质量,本文提出了一种基于Elman网络和PD复合控制的自适应逆控制算法。通过对Elman网络模型和控制对象的分析,设计了独立的指令跟踪回路和干扰抑制回路,并将逆控制和PD复合控制思想应用在干扰抑制回路中,实现了Elman网络在线学习和对被控对象的在线辨识。仿真实验结果表明,该方法能有效克服系统慢时变、干扰等非线性因素的影响,增强系统的鲁棒性
Elman Ağının Benzetilmiş Tavlama Algoritması Kullanarak Eğitilmesi
No full textYaygın olarak kullanılan geribeslemeli yapay sinir ağlarından birisi Elman ağıdır. Son yıllarda Elman ağı ve geliştirilmiş modelleri sistem kimliklendirme uygulamalarında sıkça kullanılmaktadır. Orijinal Elman ağı ve geliştirilmiş modelleri ileribesleme ve geribesleme bağlantılarına sahiptir. Ancak, bu ağlar temelde ileribeslemeli ağlar gibi standart geriyayılım algoritması ile eğitilmekte, geribesleme bağlantıları ise sabit kalmaktadır. Eğitme başarısı için, geribesleme bağlantılarının doğru değerde seçilmesi önemlidir. Bununla beraber, bu değerler uzunca bir deneme yanılma işlemiyle belirlenmektedir. Bu makalede benzetilmiş tavlama algoritmasının sistem kimliklendirme amacıyla Elman ağını eğitmede kullanılması tanımlanmıştır. Benzetilmiş Tavlama algoritması, ileribesleme ve geribesleme bağlantılarının her ikisi için optimal ağırlık değerlerini sağlayabilecek, etkili bir rasgele araştırma algoritmasıdır.One of the common used recurrent neural networks is the Elman network. Recently, Elman network and it’s modified models have been used in applications of system identification. The original Elman network and it’s modified models have feedforward and feedback connections. However, so that it can be trained essentially as feedforward networks by means of the basic backpropagation algorithm, but their feedback connections have stayed as constant. For training success, it is important to select correct values for the feedback connections. However, finding these values manually can be a lengthy trial-and-error process. This paper investigates the use of simulated annealing (SA) algorithm to train the Elman network for linear and nonlinear dynamic systems identification. The SA algorithm is an efficient random search procedure which can simultaneously obtain the optimal weight values of both the feedforward and feedback connections
Generalization of Elman Networks
No full textHammer B. Generalization of Elman Networks. In: Artificial Neural Networks - ICANN '97, 7th International Conference. Proceedings. Lecture Notes in Computer Science, 1327. Berlin: Springer; 1997: 409-414
Systems Modeling Using Deep Elman Neural Network
No full textIn this paper, the modeling of complex systems using deep Elman neural network architecture is improved. The emphasis is to retrieve better deep Elman structure that emulates perfectly such dynamic systems. To achieve this goal, sigmoid activation functions in the hidden and output layer nodes are chosen and data files on considered systems for modeling and validation steps are given. Simulation results prove the ability and the efficiency of a deep Elman neural network with two hidden layers in this task
Behavioural pattern identification and prediction in intelligent environments
No full textIn this paper, the application of soft computing techniques in prediction of an occupant's behaviour in an inhabited intelligent environment is addressed. In this research, daily activities of elderly people who live in their own homes suffering from dementia are studied. Occupancy sensors are used to extract the movement patterns of the occupant. The occupancy data is then converted into temporal sequences of activities which are eventually used to predict the occupant behaviour. To build the prediction model, different dynamic recurrent neural networks are investigated. Recurrent neural networks have shown a great ability in finding the temporal relationships of input patterns. The experimental results show that non-linear autoregressive network with exogenous inputs model correctly extracts the long term prediction patterns of the occupant and outperformed the Elman network. The results presented here are validated using data generated from a simulator and real environments
And The Angels Sing
No full textPhotograph of Johnny Mercer; Photograph of Ziggie Elman; Photograph of Benny Goodmanhttps://scholarsjunction.msstate.edu/cht-sheet-music/11689/thumbnail.jp
- …
