1) value iteration
值迭代
1.
Tradional algorithms to solve POMPDs are value iteration algorithm and policy iteration algorithm.
介绍了部分可观察Markov决策过程的基本原理和决策过程,提出一种基于策略迭代和值迭代的部分可观察Markov决策算法,该算法利用线性规划和动态规划的思想,解决当信念状态空间较大时出现的"维数灾"问题,得到Markov决策的逼近最优解。
2.
Point-Based value iteration algorithms are effective proximate algorithms to solve POMDP problems.
基于点的值迭代算法是求解POMDP问题的一类近似解法。
2) iterative threshold
迭代阈值
1.
Aiming at problems in current study of sonar image enhancement,a new enhancement method based iterative threshold was presented,the criterion function and frequency response were used to evaluate and compare current image enhancement arithmeti.
针对声呐图像增强算法研究较少等问题,提出了一种迭代阈值图像增强算法,并采用评价准则算法和频域响应法对常用的几种图像增强算法进行比较分析。
2.
A modification to Linearly Decreasing Weight strategy in standard Particle Swarm Optimization is taken by introducing descending index and iterative threshold.
引入递减指数和迭代阈值对基本粒子群算法中线性递减权策略进行了改进,在优化迭代过程中,惯性权重随当前迭代次数、指数递减率和迭代阈值非线性变化。
3.
The identification card scanned image is seriously affected by laser restraining false shadow grid, several binarization methods of document image are studied according to this characteristic, an image binarization method is proposed , which is called iterative threshold methodbased on the neighborhood feature of the pixels .
针对身份证扫描图像受激光防伪阴影网格线影响较大的特点,对图像二值化的方法进行了深入分析和研究,提出了一种基于像素邻域特征的迭代阈值方法。
3) numerical iteration
数值迭代
1.
Using the method of discretization and numerical iteration, the precise visualization results of the distributions of electric field and charges are obtained by Matlab numerical simulation.
利用离散化与数值迭代的方法,通过Matlab数值仿真得出了其电场和电荷分布的较精确的可视化结果。
2.
With the numeric calculating method applied to the reservoir flood routing operation,we introduced the numerical iteration,a new method which makes porgramme designing for flood routing operation easier.
将数值算法应用于水库防洪调算 ,提出一种适宜计算机程序设计的调洪演算新方法———数值迭代法 。
4) increment of sucessive approximation
迭代增值
5) initial value
迭代初值
1.
DTIE could be employed to provide the iterative initial value implementing M-step in some EM algorithms.
为了解决这一问题,本文改变BILOG的估计方案中迭代算法的初值,即建议使用本文给出的双重两步迭代估计(DTIE)作为迭代初值,所得新的估计方案记为MMLE/EM(DTIE)。
6) value iteration
数值迭代
1.
By the equivalent Markov process, formulas of performance potentials and average-cost optimality equations for SMCPs are derived, and a policy iteration algorithm and a value iteration algorithm are proposed, which can lead to an optimal or suboptimal stationary policy in a finite number of iterations.
利用等价Markov过程的方法,导出了SMCP的性能势公式和平均代价最优性方程,给出了求解最优或次最优平稳策略的策略迭代算法和数值迭代算法,并证明了算法的收敛性。
2.
A fast value iteration algorithm, which leads to an ε optimal stationary policy, is proposed and the convergence of thi.
文章采用无穷小生成元和性能势的基本性质 ,直接导出了平均代价模型在紧致行动集上的最优性方程及其解的存在性定理 ,提出了求解ε 最优平稳控制策略的数值迭代算法 ,并给出了这种算法的收敛性证明 。
补充资料：层层迭迭
1.见"层层迭迭"。
说明：补充资料仅用于学习参考，请勿用于其它任何用途。
参考词条