这些算法会在用户每次手动调整温度时进行自我学习。
Those algorithms refine themselves every time you manually adjust the temperature.
然后通过梯度下降法和最小二乘法相结合的混合学习算法，对控制器参数进行调整以提高其控制精度。
Then some parameters of the controller are modulated by hybrid learning algorithm of ladder descent (LD) and least square error (LSE) so as to attain better control precision.
针对BP神经网络的缺点，研究了一种动态自适应调整学习参数的改进型BP算法。
To BP neural shortcoming of network, study one dynamic self-adaptation is it study improvement type BP algorithm of parameter to adjust.
讨论了误差逆传播算法存在的缺陷，并针对其缺陷提出了动态调整学习因子与合理选取激发函数相结合的改进方案。
Disadvantages of the back propagation algorithm are discussed, and the improved methods based on dynamic learning constants and different activation functions are presented.
使用自适应学习率的算法调整网络的权值，加快了网络的学习速度。
Using an algorithms of adaptive learning rates adjust the network's weight for quickening learning rates.
该算法能够删除掉冗余的连接甚至节点，通过对网络学习步长的动态调整，避免了算法收敛速度过慢和反复震荡的问题。
The algorithm can remove redundant link even nodes on the network, through the network learning step dynamic adjustment to avoid convergence speed of.
本章重点从邻域函数、学习率调整等方面研究了二维网络的改进算法，并将之应用于烟叶动态分类问题。
The algorithm of two-dimensional Kohonen network is improved from serval aspects such as neighborhood function, learning rate, etc. It is applied into tobacco clustering.
新控制器在控制过程中借助模糊神经网络的自学习算法实现控制参数的在线调整。
The parameters of new controller can be adjusted on line based on the ability of fuzzy ne ural network.
此外，退火的鲁棒学习算法，提出这些隐藏的节点参数以及SVR - NN的权重的调整。
Furthermore, an annealing robust learning algorithm is presented to adjust these hidden node parameters as well as the weights of the SVR-NN.
随后，动态优化学习参数算法动态地调整和选取优化的学习参数。
The dynamic optimization of learning parameters can adjust learning parameters dynamically and select optimal learning parameters.
在控制算法中提出了自调整学习速率和学习初期的分层控制方法。
In the control algorithm, methods of multiple level control during initial stage of studying with the adaptive learning rate are put forward.
本文提出了一种改进的BP算法，该算法基于黄金分割法自适应调整网络学习速率。
This paper presents an improved BP algorithm, which can adapt learning rate using gold-segmentation.
为了提高网络的分类效果以及训练速度，采用了附加动量法和自适应学习速率调整法对BP算法进行了改进。
To improve the networks'classification effect and train speed, the additive momentum and self-adaptive–study-rate adjustment method are adopted further to improve traditional BP algorithm.
学习过程中，采用无监督学习算法对输入权重进行调整，采用有监督学习算法对输出权重进行调整。
Unsupervised learning is used to adjust input weight values and supervised learning is utilized to adjust output weight values.
应用一种变结构神经网络算法对初始化的模糊规则进行调整，提高模糊控制规则的自学习和自适应能力。
A kind of variable structure neural network algorithm is adopted to adjust fuzzy rules, and improves the ability of self-studying and self-adjusting in fuzzy control rules.
这种自适应模糊控制器基于模糊推理规则自学习和自调整的控制算法，无需知道太多的专家控制规则，因此解决了制冷系统MIMO模糊推理规则难以获取的问题。
This adaptive fuzzy controller is based on fuzzy inference rules self-learning without needing so much expert control rules, which solves the problem of acquiring MIMO fuzzy inference rules.
该自组织BP网络算法能够根据当前收敛状态自动调整学习率，使得网络收敛速度与学习率变化保持一致。
The algorithm can change network's learning rate followed by network's convergence state, and can adjust network's structure based on the neurons' change and their relationship.
采用一种修正恒模算法(MCMA)，该算法使修正的误差函数最小并且自适应学习率由接收序列即时调整。
In the paper, a modified constant modulus algorithm (MCMA) is proposed. The proposed algorithm minimizes a modified error function and the learning-rate is multiplied by received sequences.
其主要是令机械臂利用实践及结果通过神经网络算法来控制并调整自身行为从而不断趋向于完成某一目标，进而达到“学习”的目的。
As with Go, the skills required to have a robot manipulate objects, or perform other tasks, can be complex to program by hand.
其主要是令机械臂利用实践及结果通过神经网络算法来控制并调整自身行为从而不断趋向于完成某一目标，进而达到“学习”的目的。
As with Go, the skills required to have a robot manipulate objects, or perform other tasks, can be complex to program by hand.
应用推荐