# Alan Weiss

### MathWorks

I write documentation for MATLAB mathematical toolboxes, primarily optimization. I have also written documentation for PDE, statistics, symbolic math, and econometrics.

My pre-MathWorks job was with Bell Labs, primarily in mathematical models of data traffic, with a strong interest in parallel computation and in rare events (large deviations).

I do not provide private consulting. If you have a question, please ask on a public forum such as MATLAB Answers.

**Statistics**

RANK**35**

of 262,960

REPUTATION**5,154**

CONTRIBUTIONS

**0** Questions

**2,323** Answers

ANSWER ACCEPTANCE **
0.00%
**

VOTES RECEIVED**641**

RANK**4,855** of 17,997

REPUTATION**244**

AVERAGE RATING**0.00**

CONTRIBUTIONS**3** Files

DOWNLOADS **3**

ALL TIME DOWNLOADS**2448**

CONTRIBUTIONS**0 Posts**

CONTRIBUTIONS**0** Public Channels

AVERAGE RATING

CONTRIBUTIONS**0** Highlights

AVERAGE NO. OF LIKES

**Content Feed**

Supplying gradient for the subset of parameters

The short answer is no: there is no provision to provide analytic gradients for a subset of the parameters. I cannot think of a ...

7 days ago | 0

| accepted

Is there a way to specify objective and constraint gradients when doing nonlinear optimization in the problem-based framework?

The problem-based approach calculates and uses gradients automatically for supported functions. However, to include gradients of...

9 days ago | 0

| accepted

surrogateopt: output function

Try this. lb = [-3 -3]; ub = -lb; opts = optimoptions('surrogateopt','OutputFcn',@listout); obj = @(x)100*(x(2) - x(1)^2)^2 ...

9 days ago | 1

| accepted

Passing values to PSO options

It works for me. Here is a little test script: fun = @(x)x(1)*exp(-norm(x)^2); lb = [-5,-5]; ub = -lb; opts = optimoptions('...

9 days ago | 1

Do I need to normalize input parameters to Particle Swarm Optimization?

You don't usually need to scale your parameters, but many problems are solved more reliably if you do. See Center and Scale Your...

10 days ago | 0

Possible bug with coneprog?

I think that you have identified a problem with the default options for trajectory problems: the optimality tolerance should be ...

13 days ago | 0

| accepted

How can I save the output parameters for each step using lsqnonlin function?

Write an output function to save the values, either to a workspace variable or to a file. For syntax details, see Output Functio...

15 days ago | 0

| accepted

I'm trying to apply conditional constraints on two variable with bayesopt

I think that you set the wrong name-value argument for your constraint function. bayesopt thinks that you are using an XConstrai...

16 days ago | 0

parameter optimization in function

Try putting the following code at the end of yours: [dfinal,resnorm] = fminbnd(@(damping)trytofit(damping,x0,Acc_clean),0.00002...

27 days ago | 0

Conversion from symbolic to matlabFunction and writing the result to a file

Would it be possible to do the following: create a multiplier variable m in addition to the other variables? syms Dm0 Ds0 Ds1 d...

28 days ago | 0

Using Multistart in an unconstrained optimization

This type of problem is well-known to be numerically touchy, and to have multiple local minima. You really should give bounds, t...

29 days ago | 0

| accepted

How does using same variables for both objective and constraints affect optimization results?

By changing the objective you also change the steps that fmincon takes to solve your problem. You should not be surprised that t...

1 month ago | 0

Power balance in a microgrid

Usually the way to include this type of equality constraint is to use the approach shown in Integer and Logical Modeling. For an...

1 month ago | 0

GA doesn't proceed to the next generation

Thank you for providing a complete problem, including all supportng function files. My main question is, why are you trying to ...

1 month ago | 0

consider Z = x*y, a highly nonlinear relationship in MILP

Usually, when you have a MILP then you don't have variables multiply each other unless one of the variables is binary. For this ...

1 month ago | 0

minimize linear objective function with quadratic constraint

Again, I am not sure that I understand you correctly. I do not know what your optimization variables are, and while you say that...

1 month ago | 0

coneprog: Infeasibilities are satisfied in last step from iterative display but exitflag is still -2 (NoFeasiblePointFound)

I think that you have a misunderstanding about what the iterative display shows. For example, consider the documentation example...

1 month ago | 0

| accepted

Changing optimization technique for Gaussian process regression model

I answered a similar question recently: https://www.mathworks.com/matlabcentral/answers/1751170-svm-and-knn-hyperparameter Alan...

1 month ago | 0

| accepted

Not reproducibility in Random Forest with Hyperparameters

In creating your template for training, did you set the Reproducible name-value argument to true? Also, for your parallel option...

1 month ago | 1

| accepted

I want to solve this problem

You can use secondordercone by making a new variable , a linear objective , and another second-order cone constraint: Minimize ...

2 months ago | 1

SVM and KNN hyperparameter

Is this what you are looking for? fitcsvm hyperparameter options fitcknn hyperparameter options Alan Weiss MATLAB mathematic...

2 months ago | 0

minimize linear objective function with quadratic constraint

I'm not sure that I understand you correctly. If is an array, then perhaps you are looking for z = zeros(size(y - yhat)); m1 ...

2 months ago | 0

SVM training with hyperparameter "CVPartition"

I think that you have a misunderstanding about what these options do. Look in the first paragraph of the documentation of Hyperp...

2 months ago | 0

| accepted

Genetic Algorithm - Your fitness function must return a scalar value solution?

I am not sure what you are trying to do. Your objective function has 661 components. Are you trying to solve for 661 different v...

2 months ago | 0

| accepted

how can I automaticlly optimize hyperparameter for lasso regression. I could not set the function input as variable.

As stated in the documentation, you cannot set Alpha = 0. The Alpha parameter must be strictly positive (and no more than 1). A...

3 months ago | 0

| accepted

What variable type is "Population" in a Genetic Algorthim and how to store individuals with long genomes?

Yes, it is possible to use ga with a custom population. For an example see Custom Data Type Optimization Using the Genetic Algor...

3 months ago | 0

| accepted

Simulink using with fmincon

Perhaps you can look at some worked examples and figure it out: Using fminimax with a Simulink Model lsqnonlin with a Simulink...

3 months ago | 0

I use fcn2optimexpr,but I got an error for scalar optimizationexpression

You had a misunderstanding. The A variable should be a scalar, not a vector. All issues stemmed from that one misunderstanding. ...

3 months ago | 0

| accepted

how can I optimize a neural network with multiple outputs using the Genetic Algorithm ?

Generally speaking, optimization algorithms try to minimize a single scalar function of the input variables . I don't know what...

3 months ago | 1

how can i validate my optimization algorithm?

Your objective function is smooth. Therefore, you should not be using particleswarm or ga or simulannealbnd, but instead should ...

3 months ago | 0