Answered

How to find probability of classification in boosted tree (AdaboostM2)

For AdaBoostM1 you can convert scores into probabilities by assigning string 'doublelogit' to the ScoreTransform property of the...

How to find probability of classification in boosted tree (AdaboostM2)

For AdaBoostM1 you can convert scores into probabilities by assigning string 'doublelogit' to the ScoreTransform property of the...

5 years ago | 0

Answered

Uniform class probabilities vs. Empirical class probabilities

If you are still looking for an answer, there is only one definition for error. In each case, you form a confusion matrix and th...

Uniform class probabilities vs. Empirical class probabilities

If you are still looking for an answer, there is only one definition for error. In each case, you form a confusion matrix and th...

5 years ago | 1

| accepted

Answered

Why is LASSO in MATLAB so slow in the case of highly correlated predictors?

There could be many reasons. The |lasso| function has a lot of flexibility, so make sure you are comparing apples and apples. To...

Why is LASSO in MATLAB so slow in the case of highly correlated predictors?

There could be many reasons. The |lasso| function has a lot of flexibility, so make sure you are comparing apples and apples. To...

5 years ago | 0

| accepted

Answered

Why SVM is not giving expected result

This is a difficult problem for SVM. SVM performs best when two classes are separable or have a modest overlap. This is not the ...

Why SVM is not giving expected result

This is a difficult problem for SVM. SVM performs best when two classes are separable or have a modest overlap. This is not the ...

5 years ago | 1

| accepted

Answered

What prediction model is used in MATLAB regression tree nodes / leaves?

The prediction of a regression tree is the mean of observed responses over observations landing on this node. If you passed in o...

What prediction model is used in MATLAB regression tree nodes / leaves?

The prediction of a regression tree is the mean of observed responses over observations landing on this node. If you passed in o...

5 years ago | 0

| accepted

Answered

Fit - Nonlinear Regression - can this run faster? Python >20x faster?

I am no expert on image processing, but based on your code samples, you don't need to use |fitnlm|. When you do this x = im...

Fit - Nonlinear Regression - can this run faster? Python >20x faster?

I am no expert on image processing, but based on your code samples, you don't need to use |fitnlm|. When you do this x = im...

5 years ago | 0

| accepted

Answered

How do I use the new Support Vector Machine Regression model to simulate the response of new predictors?

This likely means that some variables in the new data have values well outside their ranges in the training data. Think about wh...

How do I use the new Support Vector Machine Regression model to simulate the response of new predictors?

This likely means that some variables in the new data have values well outside their ranges in the training data. Think about wh...

5 years ago | 0

| accepted

Answered

SVM (fitcecoc): norm(Mdl.BinaryLearners{1}.Beta) does not equal 1

You need to read the whole section and the one that follows, Computing the support vector classifier. If you, you will notice th...

SVM (fitcecoc): norm(Mdl.BinaryLearners{1}.Beta) does not equal 1

You need to read the whole section and the one that follows, Computing the support vector classifier. If you, you will notice th...

5 years ago | 0

| accepted

Answered

GPU computing for machine learning (bagging / ensemble)

There is no GPU support for decision trees or their ensembles. If you work in a sufficiently recent release, decision trees are ...

GPU computing for machine learning (bagging / ensemble)

There is no GPU support for decision trees or their ensembles. If you work in a sufficiently recent release, decision trees are ...

5 years ago | 0

Answered

How to access training data in regression trees in TreeBagger

Logical indices of observations used for each tree are stored in the |OOBIndices| property. This property wouldn't tell you thou...

How to access training data in regression trees in TreeBagger

Logical indices of observations used for each tree are stored in the |OOBIndices| property. This property wouldn't tell you thou...

5 years ago | 0

Answered

SVM with Dummy Variables

Just convert your cell array into a matrix. Yes, dummy variables will lose their identity in the sense that different levels of ...

SVM with Dummy Variables

Just convert your cell array into a matrix. Yes, dummy variables will lose their identity in the sense that different levels of ...

5 years ago | 0

| accepted

Answered

Defining cost function in partitioned SVM model

_So, my question is: does Matlab update the the priors and reset the cost function to the default as an equivalent operation to ...

Defining cost function in partitioned SVM model

_So, my question is: does Matlab update the the priors and reset the cost function to the default as an equivalent operation to ...

5 years ago | 1

Answered

Score calculation in ClassificationSVM using linear kernel function

In 14a and 14b, the Beta coefficients of an SVM model need to be divided by KernelParameters.Scale to get correct predictions. I...

Score calculation in ClassificationSVM using linear kernel function

In 14a and 14b, the Beta coefficients of an SVM model need to be divided by KernelParameters.Scale to get correct predictions. I...

5 years ago | 1

| accepted

Answered

SVM: How is the classification error with leave-one-out cross validation calculated?

|fitcsvm| passes class prior probabilities found *from the entire data* into each fold. Look at |CVSVMModel.Trained{1}.Prior|, |...

SVM: How is the classification error with leave-one-out cross validation calculated?

|fitcsvm| passes class prior probabilities found *from the entire data* into each fold. Look at |CVSVMModel.Trained{1}.Prior|, |...

5 years ago | 0

| accepted

Answered

Classification / Regression Trees for multiple responses?

Suppose your predictors are arranged in matrix X of size N-by-P and your responses are arranged in matrix Y of size N-by-R. Then...

Classification / Regression Trees for multiple responses?

Suppose your predictors are arranged in matrix X of size N-by-P and your responses are arranged in matrix Y of size N-by-R. Then...

5 years ago | 0

Answered

Save Regression Tree using Matlab commands

You can choose SaveAs from the figure pop-up menu when you click on File in the upper left corner. If you can't find SaveAs f...

Save Regression Tree using Matlab commands

You can choose SaveAs from the figure pop-up menu when you click on File in the upper left corner. If you can't find SaveAs f...

5 years ago | 0

Answered

Differences between svmtrain and fitcsvm

Look at the doc/help for |fitcsvm| or, alternatively look at the |ConvergenceInfo| property in the returned object. There are se...

Differences between svmtrain and fitcsvm

Look at the doc/help for |fitcsvm| or, alternatively look at the |ConvergenceInfo| property in the returned object. There are se...

5 years ago | 1

| accepted

Answered

svmtrain compared to fitcsvm

Check the |ConvergenceInfo.Converged| property of the returned object to see if optimization converged. This behavior of |fit...

svmtrain compared to fitcsvm

Check the |ConvergenceInfo.Converged| property of the returned object to see if optimization converged. This behavior of |fit...

5 years ago | 3

| accepted

Answered

How to set Sigma value in CompactClassificationSVM in Matlab

Binary SVM models trained by ECOC are saved in the BinaryLearners property, which is a cell array. You can't change the Sigma pr...

How to set Sigma value in CompactClassificationSVM in Matlab

Binary SVM models trained by ECOC are saved in the BinaryLearners property, which is a cell array. You can't change the Sigma pr...

5 years ago | 1

| accepted

Answered

How can I train SVM in Matlab, with more than 2 classes?

|multisvm| appears to be built on top of the older, slower |svmtrain| function, while |fitcecoc| uses the newer, faster C++ impl...

How can I train SVM in Matlab, with more than 2 classes?

|multisvm| appears to be built on top of the older, slower |svmtrain| function, while |fitcecoc| uses the newer, faster C++ impl...

5 years ago | 1

| accepted

Answered

How to input scores to the predict function

The error message says "You must pass scores as a vector..." You are passing it as a matrix with two columns. How would |perfcur...

How to input scores to the predict function

The error message says "You must pass scores as a vector..." You are passing it as a matrix with two columns. How would |perfcur...

5 years ago | 1

| accepted

Answered

how to transform output SVM into probability?

If you have the Statistics and Machine Learning Toolbox in a sufficiently recent MATLAB release, use the |fitcecoc| function and...

how to transform output SVM into probability?

If you have the Statistics and Machine Learning Toolbox in a sufficiently recent MATLAB release, use the |fitcecoc| function and...

5 years ago | 0

Answered

Cross-validation ROC curve in discriminant analysis

In principle, I don't see anything wrong with your proposal 1. There is a caveat however. You shouldn't use the same data to obt...

Cross-validation ROC curve in discriminant analysis

In principle, I don't see anything wrong with your proposal 1. There is a caveat however. You shouldn't use the same data to obt...

5 years ago | 0

Answered

How do I get the "depth" of a tree?

This should work: function depth = treedepth(tree) parent = tree.Parent; depth = 0; node = parent(end); ...

How do I get the "depth" of a tree?

This should work: function depth = treedepth(tree) parent = tree.Parent; depth = 0; node = parent(end); ...

5 years ago | 1

| accepted

Answered

ROC analysis and perfcurve

How about putting a minus sign in front of the scores (2nd input to |perfcurve|)?

ROC analysis and perfcurve

How about putting a minus sign in front of the scores (2nd input to |perfcurve|)?

5 years ago | 0

Answered

LDA: ordinal vs. categorical response

There is a difference. This choice, ordinal vs nominal, should be driven by what your response variable represents. If you ha...

LDA: ordinal vs. categorical response

There is a difference. This choice, ordinal vs nominal, should be driven by what your response variable represents. If you ha...

5 years ago | 0

Answered

Is it possible to set specific OOBIndices in TreeBagger

To train TreeBagger, you pass X and Y. If "image" is your entire data X and if "areas" are your class labels Y, you already have...

Is it possible to set specific OOBIndices in TreeBagger

To train TreeBagger, you pass X and Y. If "image" is your entire data X and if "areas" are your class labels Y, you already have...

6 years ago | 0

Answered

LDA transformation matrix for discriminative feature extraction

I don't know how you define W (when I click on the link, I get a 404 error not found). But I can make an observation: A correctl...

LDA transformation matrix for discriminative feature extraction

I don't know how you define W (when I click on the link, I get a 404 error not found). But I can make an observation: A correctl...

6 years ago | 0

Answered

Why does TreeBagger in Matlab 2014a/b only use few workers from a parallel pool?

TreeBagger does not limit the number of used cores in any way. Everything is set by your parpool configuration. The answer m...

Why does TreeBagger in Matlab 2014a/b only use few workers from a parallel pool?

TreeBagger does not limit the number of used cores in any way. Everything is set by your parpool configuration. The answer m...

6 years ago | 0

Answered

SVM convergence info objective function value - can I use it for feature selection?

There is no paper attached, and even if there was, it's unlikely I would read it. What you want to use as an indicator of the fe...

SVM convergence info objective function value - can I use it for feature selection?

There is no paper attached, and even if there was, it's unlikely I would read it. What you want to use as an indicator of the fe...

6 years ago | 0