initial weight造句
例句與造句
- Application of k - l transformation the optimization of initial weights of bp neural network
網(wǎng)絡(luò)初始權(quán)值優(yōu)化中的應(yīng)用 - Nntcs employs genetic algorithm ( ga ) in the stage of training to optimize initial weights of ann
訓(xùn)練過程中結(jié)合遺傳算法，優(yōu)化神經(jīng)網(wǎng)絡(luò)的初始權(quán)值。 - Genetic algorithm is used to optimize the initial weight of back propagation network and the operation efficiency is enhanced
用遺傳算法優(yōu)化bp網(wǎng)絡(luò)的初始權(quán)值,提高神經(jīng)網(wǎng)絡(luò)的運(yùn)算速度。 - The initial weights of the neural network can be given according to the material meaning , which expedites the network convergence
文中將神經(jīng)網(wǎng)絡(luò)與ip控制器結(jié)合，權(quán)的初始值可據(jù)其意義設(shè)定，大大加快了網(wǎng)絡(luò)的收斂速度。 - In the control process uses two bp network . one is used as nni recognizing the model , another as neural network control device ( nnc ) . but first off - line recognizes controlled device , make sure nnc initial weights
在控制的過程中,采用兩個(gè)bp網(wǎng)絡(luò),一個(gè)作為神經(jīng)網(wǎng)絡(luò)辨識器( nni )進(jìn)行辨識建模;另一個(gè)作為神經(jīng)網(wǎng)絡(luò)控制器( nnc ) 。 - It's difficult to find initial weight in a sentence. 用initial weight造句挺難的
- The algorithms for training weights update and constructing the target vectors are discussed . use the penalty term to improve the astringency of network . and study how choice the appropriate initial weights
著重研究了根據(jù)輸入和輸出量合理選擇網(wǎng)絡(luò)結(jié)構(gòu),訓(xùn)練權(quán)值的更新算法,目標(biāo)向量的合理構(gòu)造,帶懲罰項(xiàng)的bp網(wǎng)絡(luò),改善了網(wǎng)絡(luò)的收斂性。 - The dependences in multitemporal multispectral images by independent component analysis are reduced . in the algorithm , damped factor is imported to reduce the dependence on initial weights , thus the robust of the algorithm is improved
在改進(jìn)的獨(dú)立成分學(xué)習(xí)算法中,通過在梯度下降方法中引入阻尼因子,降低了對初始值的依賴,提高了獨(dú)立成分求解的穩(wěn)健性。 - During the course of develop fault diagnostic method , the influence to the training circle number with network structure 、 learning rate 、 initial weight value & door value etc are discussed . by comprehensive analyses and comparing , the comparatively rational value is adopted to be network ' s eigenvalue
在制粉系統(tǒng)故障診斷樣本訓(xùn)練過程中，本文作者探討了網(wǎng)絡(luò)結(jié)構(gòu)、學(xué)習(xí)率、初始權(quán)值閾值等因素對訓(xùn)練速度的影響，為選取合理的網(wǎng)絡(luò)參數(shù)提供了依據(jù)。 - ( 4 ) research on ann model joined with ga for area rainfall forecast the method is taken to join the genetic algorithm ( ga ) and bp algorithm together and supplementing mutually by optimizing the initial weights of ann with ga , and some application has been made in the binjiang basin for precipitation forecast
( 4 )建立了基于遺傳算法的降雨預(yù)報(bào)神經(jīng)網(wǎng)絡(luò)模型利用濱江流域的雨量站和周圍探空站的觀測資料，首次將遺傳算法( ga )應(yīng)用于流域面降雨量預(yù)報(bào)研究。 - Sofm neural networks is embedded into evolutionary strategy ( es ) . fitness function is constructed based on the state of sofm neural networks . the sensitivity of sofm neural networks to initial weight matrix and sequence of input exemplars is overcome by the strong global optimum of es
將sofm網(wǎng)絡(luò)嵌入到進(jìn)化策略（ es ）中，根據(jù)sofm網(wǎng)絡(luò)的運(yùn)行狀態(tài)構(gòu)造es的適應(yīng)性函數(shù)，利用es的強(qiáng)搜索能力，克服sofm網(wǎng)絡(luò)聚類效果受輸入模式次序和網(wǎng)絡(luò)初始連接權(quán)矩陣的影響。 - Firstly , influence factors of generalization of neural network are presented in this thesis , in order to improve neural network ’ s generalization ability and dynamic knowledge acquirement adaptive ability , a structure auto - adaptive neural network new model based on genetic algorithm is proposed to optimize structure parameter of nn including hidden layer nodes , training epochs , initial weights , and so on ; secondly , through establishing integrating neural network and introducing data fusion technique , the integrality and precision of acquired knowledge is greatly improved . then aiming at the incompleteness and uncertainty problem consisting in the process of knowledge acquirement , knowledge acquirement method based on rough sets is explored to fulfill the rule extraction for intelligent diagnosis expert system , by completing missing value data and eliminating unnecessary attributes , discretization of continuous attribute , reducing redundancy , extracting rules in this thesis . finally , rough sets theory and neural network are combined to form rnn ( rough neural network ) model for acquiring knowledge , in which rough sets theory is employed to carry out some preprocessing and neural network is acted as one role of dynamic knowledge acquirement , and rnn can improve the speed and quality of knowledge acquirement greatly
本文首先討論了影響神經(jīng)網(wǎng)絡(luò)的泛化能力的因素,提出了一種新的結(jié)構(gòu)自適應(yīng)神經(jīng)網(wǎng)絡(luò)學(xué)習(xí)算法,在新方法中,采用了遺傳算法對神經(jīng)網(wǎng)絡(luò)的結(jié)構(gòu)參數(shù)(隱層節(jié)點(diǎn)數(shù)、訓(xùn)練精度、初始權(quán)值)進(jìn)行優(yōu)化,大大提高了神經(jīng)網(wǎng)絡(luò)的泛化能力和知識動態(tài)獲取自適應(yīng)能力;其次,構(gòu)造集成神經(jīng)網(wǎng)絡(luò),引入數(shù)據(jù)融合算法,實(shí)現(xiàn)了基于集成神經(jīng)網(wǎng)絡(luò)的融合診斷,有效地提高了知識獲取的全面性、完善性及精度;然后,針對知識獲取過程中所存在的不確定性、不完備性等問題,探討了運(yùn)用粗糙集理論的知識獲取方法,通過缺損數(shù)據(jù)補(bǔ)齊、連續(xù)數(shù)據(jù)的離散、沖突消除、冗余信息約簡、知識規(guī)則抽取等一系列的算法實(shí)現(xiàn)了智能診斷的知識規(guī)則獲取;最后,將粗糙集理論與神經(jīng)網(wǎng)絡(luò)相結(jié)合,研究了粗糙集-神經(jīng)網(wǎng)絡(luò)的知識獲取方法。 - These are employed for constructing and configuring fuzzy neural network , where the number of neurons of hidden layer of network is equated to the number of rules and the initial weights of network are configured by above factors
首先利用粗糙集理論對樣本數(shù)據(jù)進(jìn)行初步規(guī)則獲取，并計(jì)算規(guī)則的依賴度和條件覆蓋度，然后根據(jù)這些規(guī)則進(jìn)行網(wǎng)絡(luò)設(shè)計(jì)，其中，網(wǎng)絡(luò)隱層節(jié)點(diǎn)的數(shù)目等于規(guī)則的數(shù)目，初始網(wǎng)絡(luò)權(quán)重由規(guī)則的依賴度和條件覆蓋度確定，最后用遺傳算法對模糊神經(jīng)網(wǎng)絡(luò)參數(shù)進(jìn)行優(yōu)化。 - The multistage constant modulus ( cm ) array is a cascade adaptive beamforming system that can recover several narrowband co - channel signals without training . the main idea of the smi - cma is to use smi to determine the initial weight for cma operation . the method can come up with the desire signal in despite of the interfering signal is stronger than the desire signal
基于以上考慮，我們提出了基于smi - cma聯(lián)合自適應(yīng)方法，該算法可以分離多個(gè)同信道信源，由smi算法決定cma算法的初始權(quán)向量，在干擾信號較強(qiáng)時(shí)，仍有穩(wěn)定的sinr輸出，具有較快的收斂速度。
更多例句： 下一頁