您要查找的是不是:
- Thus, both slow convergence and randomicity in the time-interval between samples on the genetic algorithm are overcome. 克服了遗传算法收敛慢和每个采样周期输出的随机性。
- The algorithm overcomes the drawbacks of conventional BP training such as slow convergence and the tendency to be entrapped in local minimum. 该法克服了传统BP算法因用梯度下降和误差逆向传播而拖慢收敛速度及易陷于局部极小的缺点。
- As for the slow convergence rate of BP algorithm, a momentum item was introduced into BP algorithm so that the convergence rate was increased. 针对BP神经网络收敛慢的特点,在实际算法中引入了动量项,从而提高了网络收敛速度。
- Q-learning is a typical RL method with a slow convergence speed especially as the scales of the state space and the action space increase. 利用模糊综合决策方法处理专家经验和环境信息得到Q学习的先验知识,对Q学习的初始状态进行优化。
- GA is a randomoptimization algorithm with global optimum capability, but it has the disadvantages of slow convergence and precocity. 遗传算法是一种具有全局寻优能力的随机搜索算法,但其本身存在收敛速度慢和易早熟的缺陷。
- The model overcomes some flaws of popular learning algorithm such as local minima, slow convergence and initialized values. 该模型克服了传统入侵检测系统所存在的局部极小、收敛速度慢、初值敏感性等问题。
- To overcome the default of slow convergence speed,precocity and stagnation in the basic ant colony Algorithm(ACA),we proposed an Efficient Ant Colony Algorithm(EACA). 摘要 为了克服基本蚁群算法求解速度慢、易于出现早熟和停滞现象的缺陷,提出了一种高效的蚁群算法(EACA)。
- However, the training of NNs by conventional back-propagation (BP), i. e. the BP-NNs, has intrinsic vulnerable weakness in slow convergence and local minina. 常规的神经网络权值训练算法,例如误差反传算法,存在着收敛速度慢,容易陷入局部极值点等问题。
- An improved PSO (particle swarm optimization) algorithm is presented which well addresses slow convergence speed and low calculation precision in the basic PSO algorithm. 摘要提出了一种改进的粒子群算法,很好地解决了基本粒子群算法中易陷入局部最优的缺点。
- Then some defects such as slow convergence rate and getting into local minimum in BP algorithm are pointed out,and the root of the defects is presented. 分析了BP算法的基本原理,指出了BP算法具有收敛速度慢、易陷入局部极小点等缺陷以及这些缺陷产生的根源。
- But BP network has many intrinsic defects, the structure is difficult to confirm,the blindness that initial weights is chosen results in slow convergence speed andeasily falling into local minimum. 但 bp 网络有很多固有缺陷,结构难确定,初始权值选择的盲目性导致训练速度慢,容易陷入局部最小。
- The extension of reinforcement learning to MDPs with large state,action space and high complexity has inevitably encountered the problem of the curse of dimensionality,which results in slow convergence and long training time. 传统的强化学习算法应用到大状态、动作空间和任务复杂的马尔可夫决策过程问题时;存在收敛速度慢;训练时间长等问题.
- Ant colony algorithm possesses powerful ability in searching better solutions coexisting with the disadvantages such as easily immersing into stagnation, slow convergence speed and so on. 摘要蚁群算法具有较强的发现较好解的能力,但同时也存在一些缺点,如容易出现停滞现象、收敛速度慢等。
- A multi-step Q-learning algorithm was employed to overcome the slow convergence rate of standard Q-learning, and CMAC neural network was used to generalize the continuous state space. 为解决连续过程的学习问题,采用CMAC神经网络对连续状态空间进行泛化。
- For investigation on the slow convergence speed of channel estimation and the unreason of state estimation in the probabilistic algorithms, a new multiuser detection was presented. 摘要针对以概率统计为基础信道估计收敛速度慢、状态估计存在非合理性的问题进行研究,提出了一种新的多用户检测方法。
- Simple genetic algorithm has a slow convergence velocity in late evolution and gets premature convergence easily.To solve these problems, an immune learning based genetic algorithm (ILGA) is proposed. 摘要针对基本遗传算法在进化后期收敛速度慢、易早熟收敛的问题,提出一种基于免疫学习机制的遗传算法(ILGA)。
- To solve the problems of slow convergence speed and low accuracy of the multilayer feedforward process neural networks, a double parallel feedforward process neural networks model is proposed. 摘要为克服多层前向过程神经网络收敛速度慢、精度低的问题,提出了一种双并联前向过程神经网络模型。
- Abstract: Surface electromyography is gradually used because traditional neural network has low rate of correct motion pattern recognition, and slow convergence rate of the network. 摘要: 针对传统神经网络模式识别率低、收敛速度慢等缺点,提出用支持向量机处理表面肌电信号。
- This very slowly converging series was known to Leibniz in 1674. 这个收敛很慢的级数是莱布尼茨在1674年得到的。
- Using PF to train the neural networks can overcome the drawbacks of BP algorithm such as falling into local minima and slow convergence, and the linearized error of Extended Kalman Filter. 采用粒子滤波训练神经网络克服了BP算法收敛速度慢、易陷入局部极小值的缺陷,及扩展卡尔曼滤波方法带来的模型线性化损失;