tag:www.mathworks.com,2005:/matlabcentral/fileexchange/feed
MATLAB Central File Exchange
icon.png
logo.png
MATLAB Central - File Exchange
User-contributed code library
2018-01-23T18:55:31-05:00
32082
1
100
tag:www.mathworks.com,2005:FileInfo/65813
2018-01-23T23:10:09Z
2018-01-23T23:10:09Z
Relaxed Support Vector Regression
This is the accompanying code for the paper: "Relaxed Support Vector Regression"
<p>This is the accompanying code for the paper: "Relaxed Support Vector Regression" submitted to Annals of Operations Research.</p>
Orestis Panagopoulos
http://www.mathworks.com/matlabcentral/profile/authors/11979985-orestis-panagopoulos
MATLAB 8.3 (R2014a)
IBM ILOG CPLEX
false
tag:www.mathworks.com,2005:FileInfo/65811
2018-01-23T20:47:59Z
2018-01-23T20:47:59Z
Minimizing the Himmelblau Function using Genetic Algorithm
An implementation of genetic algorithm for minimizing a multi-modal function
<p>This code implements a genetic algorithm that has:
<br />1. bit string representation of chromosomes
<br />2. fitness proportionate selection with elitism
<br />3. single- or two-point crossover
<br />4. single-point mutation
<br />5. max no. of iterations as stopping criterion</p>
<p>The outputs are:
<br />1. Plot of poorest, average, and best performance across generations
<br />2. Plot of all candidates as a moving scatter plot on the search space</p>
<p>The Himmelblau function is a 2-dimensional objective function with 4 known minima within the bounds [-6, 6]. Thus, the answer would be only one among the 4 minima every time the code is run. The code can be used to see the effect of changing the following Major Parameters to the performance:
<br />1. population size, N
<br />2. max number of iterations, G
<br />3. size of chromosome bit string, B
<br />4. number of elites in each generation, E
<br />5. crossover rate, Cr
<br />6. mutation rate, Mr</p>
Karl Ezra Pilario
http://www.mathworks.com/matlabcentral/profile/authors/8722869-karl-ezra-pilario
MATLAB 8.2 (R2013b)
false
tag:www.mathworks.com,2005:FileInfo/65809
2018-01-23T17:24:20Z
2018-01-23T17:25:56Z
On the calculation of Crowding Distance
This submission demonstrates the issues in the calculation of crowding distance
<p>Crowding distance is widely used in multi-objective optimization and is a measure of "density of solutions surrounding a particular solution in the population" [1]
<br />The procedure to determine the crowding-distance is given as following in the literature</p>
<p>"The crowding-distance computation requires sorting the population according to each objective function value in ascending order of magnitude. Thereafter, for each objective function, the boundary solutions (solutions with smallest and largest function values) are assigned an infinite distance value. All other intermediate solutions are assigned a distance value equal to the absolute normalized difference in the function values of two adjacent solutions. This calculation is continued with other objective functions. The overall crowding-distance value is calculated as the sum of individual distance values corresponding to each objective." [1]</p>
<p>Thus the crowding distance has to be unique for a given dataset. However some of the most common implementations [2,3,4] do not seem to be appropriately calculating the crowding distance. </p>
<p>In this submission, we determine the crowding distance of two datasets involving 4 points and 3 objectives. All the four points are non-dominated solutions. The only difference between dataset1 and dataset2 is that two of the rows are swapped. </p>
<p>As per the definition of crowding distance, both dataset1 and dataset2 should have the same crowding distance. However the use of the function "distancecrowding" of MATLAB yields two different results.</p>
<p>The same issue is prevalent in the code of Aravind Sheshadri [3] and also in the PLATEMO software [4]</p>
<p>References:
<br />[1] A fast and elitist multiobjective genetic algorithm: NSGA-II,
<br />IEEE Transactions on Evolutionary Computation, 6(2): 182-197
<br /><a href="http://ieeexplore.ieee.org/document/996017/">http://ieeexplore.ieee.org/document/996017/</a></p>
<p>[2] <a href="https://in.mathworks.com/help/gads/genetic-algorithm-options.html">https://in.mathworks.com/help/gads/genetic-algorithm-options.html</a></p>
<p>[3] <a href="https://in.mathworks.com/matlabcentral/fileexchange/10429-nsga-ii--a-multi-objective-optimization-algorithm">https://in.mathworks.com/matlabcentral/fileexchange/10429-nsga-ii--a-multi-objective-optimization-algorithm</a>
<br />function non_domination_sort_mod</p>
<p>[4] PlatEMO: A MATLAB Platform for Evolutionary Multi-Objective Optimization [Educational Forum], IEEE Computational Intelligence Magazine, 2017, 12(4): 73-87
<br /><a href="http://bimk.ahu.edu.cn/index.php?s=/Index/Software/index.html">http://bimk.ahu.edu.cn/index.php?s=/Index/Software/index.html</a> - PlatEMO v1.5 (2017/11/24)
<br />function CrowdingDistance.m</p>
SKS Labs
http://www.mathworks.com/matlabcentral/profile/authors/11826827-sks-labs
MATLAB 9.3 (R2017b)
10429
false
tag:www.mathworks.com,2005:FileInfo/65808
2018-01-23T16:31:42Z
2018-01-23T16:51:49Z
varspace(funx,xmin,xmax,numx,varargin)
Optimizes domain distribution x for y=f(x) for efficient shape representation
<p>When exporting complex curves with regions of interest and others without, it becomes useful to adaptivley query the domain of the function as opposed to evenly distributing the point density via linspace. When adaptivley adjusting the spacing of the representative domain, fewer points can be used as opposed to an evenly distributed domain for equivalent representation.
<br />In the example, the curve sin(x) is fairly linear about x=0 and accordingly fewer points are sample along this region whereas at x = +/- pi/2, more points are samples to account for the curvature.</p>
Alexander Laut
http://www.mathworks.com/matlabcentral/profile/authors/3625256-alexander-laut
MATLAB 9.0 (R2016a)
false
tag:www.mathworks.com,2005:FileInfo/65804
2018-01-23T13:55:41Z
2018-01-23T13:55:41Z
Appendix of Multiple Model Predictive Control of Solid Oxide Fuel Cell
Appendix of Multiple Model Predictive Control of Solid Oxide Fuel Cell
<p>Appendix of Multiple Model Predictive Control of Solid Oxide Fuel Cell.</p>
Long Wu
http://www.mathworks.com/matlabcentral/profile/authors/8668782-long-wu
MATLAB 9.1 (R2016b)
Simulink
MATLAB
false
tag:www.mathworks.com,2005:FileInfo/65801
2018-01-23T01:04:56Z
2018-01-23T01:09:24Z
Confusion Matrix for Multi-Class and Non-Binary Classification Problems
Plot a confusion matrix for a multi-class and non-binary classification problem
<p>This is the MATLAB 'plotconfusion' function adapted to multi-class and non-binary classification problem.
<br />confusion_matrix(targets,outputs) returns a confusion matrix plot for the target and output data in 'targets' and 'outputs', respectively.
<br />targets size - 1 x N
<br />outputs size - 1 x N
<br />On the confusion matrix plot, the rows correspond to the predicted class (Output Class), and the columns show the true class (Target Class). The diagonal cells show for how many (and what percentage) of the examples the trained network correctly estimates the classes of observations. That is, it shows what percentage of the true and predicted classes match. The off diagonal cells show where the classifier has made mistakes. The column on the far right of the plot shows the accuracy for each predicted class, while the row at the bottom of the plot shows the accuracy for each true class. The cell in the bottom right of the plot shows the overall accuracy.
<br />plotconfusion(targets,outputs,name) returns a confusion matrix plot with the title starting with 'name'.</p>
<p>Introduced in R2008a
<br />Requires Neural Network Toolbox</p>
David Franco
http://www.mathworks.com/matlabcentral/profile/authors/4288459-david-franco
MATLAB 9.3 (R2017b)
Neural Network Toolbox
false
tag:www.mathworks.com,2005:FileInfo/65800
2018-01-22T21:05:59Z
2018-01-22T22:49:21Z
Logistic Regression
Multiclass Logistic Regression Classification
<p>K class Logistic Regression Classification based on K binary logistic classifiers</p>
ABDUL WAHAB
http://www.mathworks.com/matlabcentral/profile/authors/8376189-abdul-wahab
MATLAB 9.2 (R2017a)
false
tag:www.mathworks.com,2005:FileInfo/65798
2018-01-22T20:47:22Z
2018-01-22T20:47:22Z
Electromagnetic Fields from a Horizontal Electric Dipole
Electromagnetic field components and polarization ellipse for a horizontal electric dipole
<p>This function computes all six of the electromagnetic field and the polarization ellipse parameters for a point or finite horizontal electric dipole source lying within a source layer (nominally, the ocean but could be an air layer if the HED is located at the bottom of the layer), overlain by an insulating halfspace, and underlain by a layered structure with variable thicknesses and electrical conductivity. Details may be found in the Hed_ell m-file within the zip bundle, and in AD Chave, Geophys. J. Int., 179, 1429-1457, 2009.</p>
Alan Chave
http://www.mathworks.com/matlabcentral/profile/authors/791238-alan-chave
MATLAB 9.2 (R2017a)
The function Hankel by the author contained on Matlab file exchange.
false
tag:www.mathworks.com,2005:FileInfo/65797
2018-01-22T20:14:21Z
2018-01-22T20:14:21Z
Hankel(order,r,f,dstruct,rerr,aerr)
Hankel Transform
<p>This function computes the Hankel transform defined as integral from 0 to inf f(x, dstruct) J sub order (x r) dx, where J sub order is a Bessel function of the first kind and order is either 0 or 1. The variable dstruct is a structure to pass data to f the (possibly) complex user provided function f. The variables rerr and aerr are the relative and absolute error passed to the Matlab function integral.
<br />The algorithm used is integration between the zero crossings of the Bessel function to obtain partial integrals, followed by their summation using the PadÃ© approximants. This approach is generally more accurate than digital filter algorithms. It also is able to handle well defined but divergent integrals such as f(x) = x.</p>
Alan Chave
http://www.mathworks.com/matlabcentral/profile/authors/791238-alan-chave
MATLAB 9.2 (R2017a)
Parallel Computing Toolbox
false
tag:www.mathworks.com,2005:FileInfo/65796
2018-01-22T19:44:20Z
2018-01-22T19:44:20Z
Mwps(xx,varargin)
Multitaper Power Spectrum, Including F-test and Reshaping
<p>This function computes the multitaper power spectrum, and optionally can return the degrees of freedom, the F-test for deterministic components, the reshaped power spectrum after line removal and the power contained in each line. The arguments have default values that can be changed by the user. Computation of the power at each frequency utilizes parfor for speed with large data sets, but can be changed to for if the Parallel Computing Toolbox is not available.</p>
Alan Chave
http://www.mathworks.com/matlabcentral/profile/authors/791238-alan-chave
MATLAB 9.2 (R2017a)
Parallel Computing Toolbox
Signal Processing Toolbox
Statistics and Machine Learning Toolbox
false