Beantwortet
FLD (Fisher Linear Discriminant)
If you have MATLAB 11b, take a look at class ClassificationDiscriminant. If you have an older version, take a look at function C...

mehr als 14 Jahre vor | 0

Beantwortet
classify
"Pooled" and "stratified" are conventional terms in the statistics literature. A pooled estimate is computed using all observati...

mehr als 14 Jahre vor | 0

| akzeptiert

Beantwortet
FISHER DISCRIMINANT FUNCTION - ROC
Open a documentation window by typing 'doc' in your session or selecting Product Help on MATLAB desktop. Then type 'roc' into th...

mehr als 14 Jahre vor | 0

Beantwortet
Haar feature based cascade classifiers and duplex flexible window
Sorry, I am not going to look at the paper you mention. FWIW, I suspect that Harr features is what you extract from an image usi...

mehr als 14 Jahre vor | 0

Beantwortet
Fisher Linear Discriminant
Fisher discriminant is provided in the Statistics Toolbox. If you have MATLAB 11b, use ClassificationDiscriminant. In an earlier...

mehr als 14 Jahre vor | 0

| akzeptiert

Beantwortet
TreeBagger gives different results depending on 'oobvarimp' being 'on' or 'off'
One way of assessing what is high and low under these circumstances would be to look at the classification error. It can be mode...

mehr als 14 Jahre vor | 0

Beantwortet
TreeBagger gives different results depending on 'oobvarimp' being 'on' or 'off'
Computing variable importance by permuting observations across every variable (that's what you get when you set oobvarimp to on)...

mehr als 14 Jahre vor | 0

| akzeptiert

Beantwortet
How is varimportance calculated?
http://www.mathworks.com/help/toolbox/stats/classregtree.varimportance.html If you split parent node 1 into child nodes 2 and...

mehr als 14 Jahre vor | 0

| akzeptiert

Beantwortet
Decision/Regression stumps using ClassificationTree.template or RegressionTree.template
You get stumps by default for any boosting algorithm from fitensemble. Or, equivalently, you can set 'minparent' to the number o...

mehr als 14 Jahre vor | 0

| akzeptiert

Beantwortet
Compare two trees generated using same classes but different attributes in each case
There may be a straightforward way, but it is up to you to define criteria for such a comparison. Here is a list of all tree met...

mehr als 14 Jahre vor | 0

| akzeptiert

Beantwortet
Classification and regression trees
Does this link help? http://www.mathworks.com/help/toolbox/stats/bsw6a5x.html#bsw6bqq If you are concerned with a specific...

mehr als 14 Jahre vor | 0

Beantwortet
Visualizing a tree
If you work in 11a or later and if you have Statistics Toolbox, you can boost trees using fitensemble function. Then you can use...

mehr als 14 Jahre vor | 0

Beantwortet
ClassificationTree with unequal costs
Yes, ClassificationTree always predicts class labels based on posterior probabilities. In that, ClassificationTree/predict devia...

mehr als 14 Jahre vor | 0

Beantwortet
Classregtree - implementing an additional constraint
If you work in release 9b or later, the function searching for the optimal split in a tree node is coded in C for speed. The sou...

mehr als 14 Jahre vor | 0

Beantwortet
KERNEL : mean integrated squared error- Bandwidth Selection
The "right" width depends on your assumptions about the fitted distribution. MATLAB does not choose the bandwidth "randomly". It...

mehr als 14 Jahre vor | 0

| akzeptiert

Beantwortet
How to make matrix a possitive definite
If you have 10935 predictors, your entire training and test data are not going to be enough to make the covariance matrix non-si...

mehr als 14 Jahre vor | 0

Beantwortet
Leraning classification with most training samples in one category
The answer depends on how you define a "fair" classifier. If the ultimate goal of your analysis is to minimize the overall class...

fast 15 Jahre vor | 2

| akzeptiert

Beantwortet
Difference between individual and cumulative oobMargin of TreeBagger
When you ask for an OOB margin from one tree, you get zero if this observation was in bag for this tree. The margin is undefined...

fast 15 Jahre vor | 0

| akzeptiert