Location:
Search - minima
Search list
Description: 开发环境:Matlab
简要说明:动量-自适应学习调整算法。在实际应用中，原始的BP算法很难胜任，因此出现了很多的改进算法。BP算法的改进主要有两种途径，一种是采用启发式学习方法，另一种则是采用更有效的优化算法。本例采用动量BP算法，来实现对网络的训练过程，动量法降低了网络对于误差曲面局部细节的敏感性，有效地抑制网络陷于局部极小。-development environment : Matlab Brief Description : Momentum-adaptive learning algorithm adjustments. In practical application, the original BP algorithm competence, resulting in a lot of improved algorithm. BP algorithm improvements There are two main ways of using a heuristic approach to learning Another is the use of a more effective method of optimization. Momentum cases using the BP algorithm to achieve the network training process, Momentum for reducing error of the network for local surface details of the sensitivity, to effectively curb the network into local minima.
Platform: |
Size: 1051 |
Author: zhangjian |
Hits:
Description: This demo nstrates the use of the reversible jump MCMC simulated annealing for neural networks. This algorithm enables us to maximise the joint posterior distribution of the network parameters and the number of basis function. It performs a global search in the joint space of the parameters and number of parameters, thereby surmounting the problem of local minima. It allows the user to choose among various model selection criteria, including AIC, BIC and MDL
Platform: |
Size: 958327 |
Author: 郭剑辉 |
Hits:
Description: 模拟退火优化算法 Simulated annealing is a simple and general algorithm for finding global minima. It operates by simulating the cooling of a (usually fictitious) physical system whose possible energies correspond to the values of the objective function being minimized. The anal- ogy works because physical systems occupy only the states with the lowest energy as the temperature is lowered to absolute zero. Simulated annealing has been developed by a wide and highly interdisciplinary com- munity and used by an even wider one. As a consequence, its techniques and results are scattered through the literature and are not easily accessible for the computer scientist, physicist, or chemist who wants to become familiar with the field. The present monograph is intended both as an introduction for the noninitiated and as a review for the more expert reader. We start by developing the subject from scratch. We then explain the methods and techniques indispensable for building state-of-the-art implementations. The physical background presented is meant to sharpen the reader's intuition for the field.
Platform: |
Size: 6738858 |
Author: luli395 |
Hits:
Description: 开发环境:Matlab
简要说明:动量-自适应学习调整算法。在实际应用中，原始的BP算法很难胜任，因此出现了很多的改进算法。BP算法的改进主要有两种途径，一种是采用启发式学习方法，另一种则是采用更有效的优化算法。本例采用动量BP算法，来实现对网络的训练过程，动量法降低了网络对于误差曲面局部细节的敏感性，有效地抑制网络陷于局部极小。-development environment : Matlab Brief Description : Momentum-adaptive learning algorithm adjustments. In practical application, the original BP algorithm competence, resulting in a lot of improved algorithm. BP algorithm improvements There are two main ways of using a heuristic approach to learning Another is the use of a more effective method of optimization. Momentum cases using the BP algorithm to achieve the network training process, Momentum for reducing error of the network for local surface details of the sensitivity, to effectively curb the network into local minima.
Platform: |
Size: 1024 |
Author: zhangjian |
Hits:
Description: This demo nstrates the use of the reversible jump MCMC simulated annealing for neural networks. This algorithm enables us to maximise the joint posterior distribution of the network parameters and the number of basis function. It performs a global search in the joint space of the parameters and number of parameters, thereby surmounting the problem of local minima. It allows the user to choose among various model selection criteria, including AIC, BIC and MDL
Platform: |
Size: 958464 |
Author: 大辉 |
Hits:
Description: 一种改进的混沌优化，便尺度混顿优化的优势在于速度快，不容易掉入局部极小-An improved chaos optimization, then mixed Dayton-scale optimization has the advantage of being fast, not easy to fall into local minima
Platform: |
Size: 2048 |
Author: xjwu |
Hits:
Description: 便尺度混沌优化，方法很简单，但是速度快，不易局部极小-Will scale chaos optimization method is simple, but fast, easy local minima
Platform: |
Size: 161792 |
Author: xjwu |
Hits:
Description: A collection of functions is presented which includes 2nd generation wavelet decomposition and reconstruction tools for images as well as functions for the computation of moment invariants. The wavelet schemes rely on the lifting scheme of Sweldens. Rectangular grids are split into quincunx grids, also known as red-black ordering. The prediction filters include Neville filters as well as a few nonlinear ones fairly capable of preserving local maxima or minima. The decomposition and reconstruction functions are called in the style of the Matlab Wavelet Toolbox. Many small and a few elaborate examples have been included, ranging from the
computation of moment invariants to
multiresolution image fusion. Please
see Contents.m for an exhaustive list.
-A collection of functions is presented which includes 2nd generation wavelet decomposition and reconstruction tools for images as well as functions for the computation of moment invariants. The wavelet schemes rely on the lifting scheme of Sweldens. Rectangular grids are split into quincunx grids, also known as red-black ordering. The prediction filters include Neville filters as well as a few nonlinear ones fairly capable of preserving local maxima or minima. The decomposition and reconstruction functions are called in the style of the Matlab Wavelet Toolbox. Many small and a few elaborate examples have been included, ranging from the computation of moment invariants to multiresolution image fusion. Please see Contents.m for an exhaustive list.
Platform: |
Size: 592896 |
Author: yuan |
Hits:
Description: 最优化中的实例，利用黄金分割法求出下单峰函数极小点-Optimization of the examples, the use of golden section method are obtained under the single-peak function minima
Platform: |
Size: 878592 |
Author: chunchen |
Hits:
Description: 利用神经网络分析方法对一个给定信号的两端进行延拓，在数据的两端各得到两个附加的极大值点和两个附加的极小值点.由此利用三次样条函数得到原始信号的上下包络线和平均包络线，实现了准确的EMD分解.
-The use of neural network analysis of a signal at the two ends of a given extension to the data obtained at the two ends of the two additional maxima and minima of two additional points. This use of cubic spline function has been the top and bottom of the original signal envelope and the average envelope to achieve an accurate decomposition of the EMD.
Platform: |
Size: 464896 |
Author: 齐磊 |
Hits:
Description: Simulated Annealing (SA) is a smart (meta)-heuristic for Optimization. Given a cost function in a large search space, SA replaces the current solution by a random "nearby" solution. The nearby solution is chosen with a probability that depends on the difference between the corresponding function values and on a global parameter T (a.k.a the temperature). T is gradually decreased during the process. The current solution changes almost randomly when T is large, but increasingly "downhill" as T goes to zero. The allowance for "uphill" moves saves the method from becoming stuck at local minima.
This approach has some similitude with Physic, where the heat causes the atoms to become unstuck from their initial positions and wander randomly through states of higher energy the slow cooling gives them more chances of finding configurations with lower internal energy than the initial one.-Simulated Annealing (SA) is a smart (meta)-heuristic for Optimization. Given a cost function in a large search space, SA replaces the current solution by a random " nearby" solution. The nearby solution is chosen with a probability that depends on the difference between the corresponding function values and on a global parameter T (aka the temperature). T is gradually decreased during the process. The current solution changes almost randomly when T is large, but increasingly " downhill" as T goes to zero. The allowance for " uphill" moves saves the method from becoming stuck at local minima. This approach has some similitude with Physic, where the heat causes the atoms to become unstuck from their initial positions and wander randomly through states of higher energy the slow cooling gives them more chances of finding configurations with lower internal energy than the initial one.
Platform: |
Size: 20480 |
Author: dingchong |
Hits:
Description: NUMERICAL OPTIMIZATION:
This steepest descent method with constant step length to find the minima of
f(x, y) = xy exp(− 2x^2 − y^2 + 0.3y)
Graphical represxentation in 5 ways of solution, simple and clear explained.-NUMERICAL OPTIMIZATION:
This is steepest descent method with constant step length to find the minima of
f(x, y) = xy exp(− 2x^2 − y^2+ 0.3y)
Graphical represxentation in 5 ways of solution, simple and clear explained.
Platform: |
Size: 1024 |
Author: venera |
Hits:
Description: Program for finding minima of functions of several variables, several methods (Powell, Polak-Riber, Pearson, etc.)
Platform: |
Size: 9216 |
Author: GnomE |
Hits:
Description: Program for finding minima of functions of several variables, several methods with GUI.
Platform: |
Size: 51200 |
Author: GnomE |
Hits:
Description: Peak detection is one of the most important time-domain functions performed in signal monitoring. Peak detection is the process of finding the locations and amplitudes of local maxima and minima in a signal that satisfies certain properties. These properties can be simple or complex. For example, requiring that a peak exceeds a certain threshold value is a simple property. However, requiring that a peak’s shape resembles that of a prototype peak is a complex property.
Platform: |
Size: 1024 |
Author: Kirill Sakhnov |
Hits:
Description: Min/Max filter，最大最小滤波器，说明看英文，只是国外的人写的代码，
-Description The filter computes the minima and/or maxima of an array over sliding window with a given size.
Multidimensional array is fully supported: running filter in 1D, 2D filter for image processing applications (erosion/dilatation), 3D and more.
This package has been implemented with a special care on the running speed: the MEX engine uses an algorithms that requires no more than three (3) comparisons per element and per dimension in all configurations.
All numerical and logical class arrays are supported.
Contributor (beside author) is Vaclav Potesil
Acknowledgements
This submission has inspired the following:
Free-knot spline approximation
MATLAB release MATLAB 7.8 (R2009a)
Other requirements MEX correctly setup NO image processing is required Download a tiff file for
Platform: |
Size: 29696 |
Author: 谢冉 |
Hits:
Description: Mathematical methods are an alternative to tackle visual perception. The central
idea behind these methods is to reformulate the visual perception components
as optimization problems where the minima of a specifically designed objective
function "solve" the task under consideration. The definition of such functions is
often an ill-posed problem since the number of variables to be recovered is much
larger than the number of constraints. Furthermore, often the optimization process
itself is ill-posed due the non-convexity of the designed function inducing the
presence of local minima
Platform: |
Size: 84992 |
Author: Shashidhar |
Hits:
Description: Noise Estimation by Minima Controlled Recursive Averaging for Robust Speech Enhancement
Platform: |
Size: 332800 |
Author: madin |
Hits:
Description: fast 1d and 2d local minima detector
Platform: |
Size: 2048 |
Author: samiio |
Hits:
Description: nonlinear optimization code can find local or global minima
Platform: |
Size: 1024 |
Author: haydar |
Hits: