resubPredict

Classify observations in multiclass error-correcting output codes (ECOC) model

Syntax

label = resubPredict(Mdl)

label = resubPredict(Mdl,Name,Value)

[label,NegLoss,PBScore] = resubPredict(___)

[label,NegLoss,PBScore,Posterior] = resubPredict(___)

Description

label = resubPredict(Mdl) returns a vector of predicted class labels (label) for the trained multiclass error-correcting output codes (ECOC) model Mdl using the predictor data stored in Mdl.X.

The software predicts the classification of an observation by assigning the observation to the class yielding the largest negated average binary loss (or, equivalently, the smallest average binary loss).

label = resubPredict(Mdl,Name,Value) returns predicted class labels with additional options specified by one or more name-value pair arguments. For example, specify the posterior probability estimation method, decoding scheme, or verbosity level.

[label,NegLoss,PBScore] = resubPredict(___) uses any of the input argument combinations in the previous syntaxes and additionally returns the negated average binary loss per class (NegLoss) for observations, and the positive-class scores (PBScore) for the observations classified by each binary learner.

[label,NegLoss,PBScore,Posterior] = resubPredict(___) additionally returns posterior class probability estimates for observations (Posterior).

To obtain posterior class probabilities, you must set 'FitPosterior',true when training the ECOC model using fitcecoc. Otherwise, resubPredict throws an error.

Examples

Predict Labels of Training Data Using ECOC Model

Load Fisher's iris data set. Specify the predictor data X, the response data Y, and the order of the classes in Y.

load fisheriris
X = meas;
Y = categorical(species);
classOrder = unique(Y);

Train an ECOC model using SVM binary classifiers. Standardize the predictors using an SVM template, and specify the class order.

t = templateSVM('Standardize',true);
Mdl = fitcecoc(X,Y,'Learners',t,'ClassNames',classOrder);

t is an SVM template object. During training, the software uses default values for empty properties in t. Mdl is a ClassificationECOC model.

Predict the labels of the training data. Print a random subset of true and predicted labels.

labels = resubPredict(Mdl);

rng(1); % For reproducibility
n = numel(Y); % Sample size
idx = randsample(n,10);
table(Y(idx),labels(idx),'VariableNames',{'TrueLabels','PredictedLabels'})

ans=10×2 table
    TrueLabels    PredictedLabels
    __________    _______________

    setosa          setosa       
    versicolor      versicolor   
    virginica       virginica    
    setosa          setosa       
    versicolor      versicolor   
    setosa          setosa       
    versicolor      versicolor   
    versicolor      versicolor   
    setosa          setosa       
    setosa          setosa       

Mdl correctly labels the observations with indices idx.

Predict Resubstitution Labels of ECOC Model Using Custom Binary Loss Function

Load Fisher's iris data set. Specify the predictor data X, the response data Y, and the order of the classes in Y.

load fisheriris
X = meas;
Y = categorical(species);
classOrder = unique(Y); % Class order

Train an ECOC model using SVM binary classifiers. Standardize the predictors using an SVM template, and specify the class order.

t = templateSVM('Standardize',true);
Mdl = fitcecoc(X,Y,'Learners',t,'ClassNames',classOrder);

t is an SVM template object. During training, the software uses default values for empty properties in t. Mdl is a ClassificationECOC model.

SVM scores are signed distances from the observation to the decision boundary. Therefore, the domain is $$(-\infty ,\infty )$$. Create a custom binary loss function that does the following:

• Map the coding design matrix (M) and positive-class classification scores (s) for each learner to the binary loss for each observation.

• Use linear loss.

• Aggregate the binary learner loss using the median.

You can create a separate function for the binary loss function, and then save it on the MATLAB® path. Or, you can specify an anonymous binary loss function. In this case, create a function handle (customBL) to an anonymous binary loss function.

customBL = @(M,s)median(1 - (M.*s),2,'omitnan')/2;

Predict labels for the training data and estimate the median binary loss per class. Print the median negative binary losses per class for a random set of 10 observations.

[label,NegLoss] = resubPredict(Mdl,'BinaryLoss',customBL);

rng(1); % For reproducibility
n = numel(Y); % Sample size
idx = randsample(n,10);
classOrder

classOrder = 3x1 categorical
     setosa 
     versicolor 
     virginica 

table(Y(idx),label(idx),NegLoss(idx,:),'VariableNames',...
    {'TrueLabel','PredictedLabel','NegLoss'})

ans=10×3 table
    TrueLabel     PredictedLabel                NegLoss            
    __________    ______________    _______________________________

    setosa          versicolor       0.1236      1.9565     -3.5801
    versicolor      versicolor      -1.0169       0.629     -1.1121
    virginica       virginica       -1.9079    -0.21788     0.62583
    setosa          versicolor      0.43825      2.2437      -4.182
    versicolor      versicolor      -1.0732       0.396    -0.82276
    setosa          versicolor      0.26653      2.1999     -3.9664
    versicolor      versicolor      -1.1233      0.6988     -1.0755
    versicolor      versicolor      -1.2708     0.51782      -0.747
    setosa          versicolor      0.35168      2.0674     -3.9191
    setosa          versicolor       0.2334      2.1882     -3.9216

The order of the columns corresponds to the elements of classOrder. The software predicts the label based on the maximum negated loss. The results indicate that the median of the linear losses might not perform as well as other losses.

Estimate Posterior Probabilities Using ECOC Classifier

Train an ECOC classifier using SVM binary learners. First predict the training-sample labels and class posterior probabilities. Then predict the maximum class posterior probability at each point in a grid. Visualize the results.

Load Fisher's iris data set. Specify the petal dimensions as the predictors and the species names as the response.

load fisheriris
X = meas(:,3:4);
Y = species;
rng(1); % For reproducibility

Create an SVM template. Standardize the predictors, and specify the Gaussian kernel.

t = templateSVM('Standardize',true,'KernelFunction','gaussian');

t is an SVM template. Most of its properties are empty. When the software trains the ECOC classifier, it sets the applicable properties to their default values.

Train the ECOC classifier using the SVM template. Transform classification scores to class posterior probabilities (which are returned by predict or resubPredict) using the 'FitPosterior' name-value pair argument. Specify the class order using the 'ClassNames' name-value pair argument. Display diagnostic messages during training by using the 'Verbose' name-value pair argument.

Mdl = fitcecoc(X,Y,'Learners',t,'FitPosterior',true,...
    'ClassNames',{'setosa','versicolor','virginica'},...
    'Verbose',2);

Training binary learner 1 (SVM) out of 3 with 50 negative and 50 positive observations.
Negative class indices: 2
Positive class indices: 1

Fitting posterior probabilities for learner 1 (SVM).
Training binary learner 2 (SVM) out of 3 with 50 negative and 50 positive observations.
Negative class indices: 3
Positive class indices: 1

Fitting posterior probabilities for learner 2 (SVM).
Training binary learner 3 (SVM) out of 3 with 50 negative and 50 positive observations.
Negative class indices: 3
Positive class indices: 2

Fitting posterior probabilities for learner 3 (SVM).

Mdl is a ClassificationECOC model. The same SVM template applies to each binary learner, but you can adjust options for each binary learner by passing in a cell vector of templates.

Predict the training-sample labels and class posterior probabilities. Display diagnostic messages during the computation of labels and class posterior probabilities by using the 'Verbose' name-value pair argument.

[label,~,~,Posterior] = resubPredict(Mdl,'Verbose',1);

Predictions from all learners have been computed.
Loss for all observations has been computed.
Computing posterior probabilities...

Mdl.BinaryLoss

ans = 
'quadratic'

The software assigns an observation to the class that yields the smallest average binary loss. Because all binary learners are computing posterior probabilities, the binary loss function is quadratic.

Display a random set of results.

idx = randsample(size(X,1),10,1);
Mdl.ClassNames

ans = 3x1 cell
    {'setosa'    }
    {'versicolor'}
    {'virginica' }

table(Y(idx),label(idx),Posterior(idx,:),...
    'VariableNames',{'TrueLabel','PredLabel','Posterior'})

ans=10×3 table
      TrueLabel         PredLabel                     Posterior               
    ______________    ______________    ______________________________________

    {'virginica' }    {'virginica' }     0.0039319     0.0039866       0.99208
    {'virginica' }    {'virginica' }      0.017066      0.018262       0.96467
    {'virginica' }    {'virginica' }      0.014947      0.015855        0.9692
    {'versicolor'}    {'versicolor'}    2.2197e-14       0.87318       0.12682
    {'setosa'    }    {'setosa'    }         0.999    0.00025091    0.00074639
    {'versicolor'}    {'virginica' }    2.2195e-14      0.059427       0.94057
    {'versicolor'}    {'versicolor'}    2.2194e-14       0.97002      0.029984
    {'setosa'    }    {'setosa'    }         0.999     0.0002499    0.00074741
    {'versicolor'}    {'versicolor'}     0.0085638       0.98259     0.0088482
    {'setosa'    }    {'setosa'    }         0.999    0.00025013    0.00074718

The columns of Posterior correspond to the class order of Mdl.ClassNames.

Define a grid of values in the observed predictor space. Predict the posterior probabilities for each instance in the grid.

xMax = max(X);
xMin = min(X);

x1Pts = linspace(xMin(1),xMax(1));
x2Pts = linspace(xMin(2),xMax(2));
[x1Grid,x2Grid] = meshgrid(x1Pts,x2Pts);

[~,~,~,PosteriorRegion] = predict(Mdl,[x1Grid(:),x2Grid(:)]);

For each coordinate on the grid, plot the maximum class posterior probability among all classes.

contourf(x1Grid,x2Grid,...
        reshape(max(PosteriorRegion,[],2),size(x1Grid,1),size(x1Grid,2)));
h = colorbar;
h.YLabel.String = 'Maximum posterior';
h.YLabel.FontSize = 15;

hold on
gh = gscatter(X(:,1),X(:,2),Y,'krk','*xd',8);
gh(2).LineWidth = 2;
gh(3).LineWidth = 2;

title('Iris Petal Measurements and Maximum Posterior')
xlabel('Petal length (cm)')
ylabel('Petal width (cm)')
axis tight
legend(gh,'Location','NorthWest')
hold off

Estimate Posterior Probabilities Using Parallel Computing

Train a multiclass ECOC model and estimate the posterior probabilities using parallel computing.

Load the arrhythmia data set. Examine the response data Y, and determine the number of classes.

load arrhythmia
Y = categorical(Y);
tabulate(Y)

  Value    Count   Percent
      1      245     54.20%
      2       44      9.73%
      3       15      3.32%
      4       15      3.32%
      5       13      2.88%
      6       25      5.53%
      7        3      0.66%
      8        2      0.44%
      9        9      1.99%
     10       50     11.06%
     14        4      0.88%
     15        5      1.11%
     16       22      4.87%

K = numel(unique(Y));

Several classes are not represented in the data, and many other classes have low relative frequencies.

Specify an ensemble learning template that uses the GentleBoost method and 50 weak classification tree learners.

t = templateEnsemble('GentleBoost',50,'Tree');

t is a template object. Most of its properties are empty ([]). The software uses default values for all empty properties during training.

Because the response variable contains many classes, specify a sparse random coding design.

rng(1); % For reproducibility
Coding = designecoc(K,'sparserandom');

Train an ECOC model using parallel computing. Specify to fit posterior probabilities.

pool = parpool;                      % Invokes workers

Starting parallel pool (parpool) using the 'local' profile ...
Connected to the parallel pool (number of workers: 6).

options = statset('UseParallel',true);
Mdl = fitcecoc(X,Y,'Learner',t,'Options',options,'Coding',Coding,...
    'FitPosterior',true);

Mdl is a ClassificationECOC model. You can access its properties using dot notation.

The pool invokes six workers, although the number of workers might vary among systems.

Estimate posterior probabilities, and display the posterior probability of being classified as not having arrhythmia (class 1) given a random subset of the training data.

[~,~,~,posterior] = resubPredict(Mdl);

n = numel(Y);
idx = randsample(n,10,1);
table(idx,Y(idx),posterior(idx,1),...
    'VariableNames',{'ObservationIndex','TrueLabel','PosteriorNoArrythmia'})

ans=10×3 table
    ObservationIndex    TrueLabel    PosteriorNoArrythmia
    ________________    _________    ____________________

           79              1                0.93436      
          248              1                0.95574      
          398              10              0.032378      
          207              1                0.97965      
          340              1                0.93656      
          206              1                0.97795      
          345              10              0.015642      
          296              2                0.13433      
          391              1                 0.9648      
          406              1                0.94861      

Input Arguments

Mdl — Full, trained multiclass ECOC model

ClassificationECOC model

Full, trained multiclass ECOC model, specified as a ClassificationECOC model trained with fitcecoc.

Name-Value Arguments

Specify optional pairs of arguments as Name1=Value1,...,NameN=ValueN, where Name is the argument name and Value is the corresponding value. Name-value arguments must appear after other arguments, but the order of the pairs does not matter.

Before R2021a, use commas to separate each name and value, and enclose Name in quotes.

Example: resubPredict(Mdl,'BinaryLoss','linear','Decoding','lossbased') specifies a linear binary learner loss function and a loss-based decoding scheme for aggregating the binary losses.

BinaryLoss — Binary learner loss function

"hamming" | "linear" | "logit" | "exponential" | "binodeviance" | "hinge" | "quadratic" | function handle

Binary learner loss function, specified as a built-in loss function name or function handle.

• This table describes the built-in functions, where y_j is the class label for a particular binary learner (in the set {–1,1,0}), s_j is the score for observation j, and g(y_j,s_j) is the binary loss formula.

  Value	Description	Score Domain	g(yj,sj)
  "binodeviance"	Binomial deviance	(–∞,∞)	log[1 + exp(–2yjsj)]/[2log(2)]
  "exponential"	Exponential	(–∞,∞)	exp(–yjsj)/2
  "hamming"	Hamming	[0,1] or (–∞,∞)	[1 – sign(yjsj)]/2
  "hinge"	Hinge	(–∞,∞)	max(0,1 – yjsj)/2
  "linear"	Linear	(–∞,∞)	(1 – yjsj)/2
  "logit"	Logistic	(–∞,∞)	log[1 + exp(–yjsj)]/[2log(2)]
  "quadratic"	Quadratic	[0,1]	[1 – yj(2sj – 1)]2/2

  The software normalizes binary losses so that the loss is 0.5 when y_j = 0. Also, the software calculates the mean binary loss for each class [1].

• For a custom binary loss function, for example customFunction, specify its function handle BinaryLoss=@customFunction.

  customFunction has this form:

  bLoss = customFunction(M,s)

  • M is the K-by-B coding matrix stored in Mdl.CodingMatrix.

  • s is the 1-by-B row vector of classification scores.

  • bLoss is the classification loss. This scalar aggregates the binary losses for every learner in a particular class. For example, you can use the mean binary loss to aggregate the loss over the learners for each class.

  • K is the number of classes.

  • B is the number of binary learners.

  For an example of passing a custom binary loss function, see Predict Test-Sample Labels of ECOC Model Using Custom Binary Loss Function.

This table identifies the default BinaryLoss value, which depends on the score ranges returned by the binary learners.

Assumption	Default Value
All binary learners are any of the following: Classification decision trees Discriminant analysis models k-nearest neighbor models Linear or kernel classification models of logistic regression learners Naive Bayes models	"quadratic"
All binary learners are SVMs or linear or kernel classification models of SVM learners.	"hinge"
All binary learners are ensembles trained by AdaboostM1 or GentleBoost.	"exponential"
All binary learners are ensembles trained by LogitBoost.	"binodeviance"
You specify to predict class posterior probabilities by setting FitPosterior=true in fitcecoc.	"quadratic"
Binary learners are heterogeneous and use different loss functions.	"hamming"

To check the default value, use dot notation to display the BinaryLoss property of the trained model at the command line.

Example: BinaryLoss="binodeviance"

Data Types: char | string | function_handle

Decoding — Decoding scheme

"lossweighted" (default) | "lossbased"

Decoding scheme that aggregates the binary losses, specified as "lossweighted" or "lossbased". For more information, see Binary Loss.

Example: Decoding="lossbased"

Data Types: char | string

NumKLInitializations — Number of random initial values

0 (default) | nonnegative integer scalar

Number of random initial values for fitting posterior probabilities by Kullback-Leibler divergence minimization, specified as the comma-separated pair consisting of 'NumKLInitializations' and a nonnegative integer scalar.

If you do not request the fourth output argument (Posterior) and set 'PosteriorMethod','kl' (the default), then the software ignores the value of NumKLInitializations.

For more details, see Posterior Estimation Using Kullback-Leibler Divergence.

Example: 'NumKLInitializations',5

Data Types: single | double

Options — Estimation options

[] (default) | structure array

Estimation options, specified as a structure array as returned by statset.

To invoke parallel computing you need a Parallel Computing Toolbox™ license.

Example: Options=statset(UseParallel=true)

Data Types: struct

PosteriorMethod — Posterior probability estimation method

'kl' (default) | 'qp'

Posterior probability estimation method, specified as the comma-separated pair consisting of 'PosteriorMethod' and 'kl' or 'qp'.

• If PosteriorMethod is 'kl', then the software estimates multiclass posterior probabilities by minimizing the Kullback-Leibler divergence between the predicted and expected posterior probabilities returned by binary learners. For details, see Posterior Estimation Using Kullback-Leibler Divergence.

• If PosteriorMethod is 'qp', then the software estimates multiclass posterior probabilities by solving a least-squares problem using quadratic programming. You need an Optimization Toolbox™ license to use this option. For details, see Posterior Estimation Using Quadratic Programming.

• If you do not request the fourth output argument (Posterior), then the software ignores the value of PosteriorMethod.

Example: 'PosteriorMethod','qp'

Verbose — Verbosity level

0 (default) | 1

Verbosity level, specified as 0 or 1. Verbose controls the number of diagnostic messages that the software displays in the Command Window.

If Verbose is 0, then the software does not display diagnostic messages. Otherwise, the software displays diagnostic messages.

Example: Verbose=1

Data Types: single | double

Output Arguments

label — Predicted class labels

categorical array | character array | logical vector | numeric vector | cell array of character vectors

Predicted class labels, returned as a categorical or character array, logical or numeric vector, or cell array of character vectors.

label has the same data type as Mdl.ClassNames and has the same number of rows as Mdl.X.

The software predicts the classification of an observation by assigning the observation to the class yielding the largest negated average binary loss (or, equivalently, the smallest average binary loss).

NegLoss — Negated average binary losses

numeric matrix

Negated average binary losses, returned as a numeric matrix. NegLoss is an n-by-K matrix, where n is the number of observations (size(Mdl.X,1)) and K is the number of unique classes (size(Mdl.ClassNames,1)).

NegLoss(i,k) is the negated average binary loss for classifying observation i into the kth class.

• If Decoding is 'lossbased', then NegLoss(i,k) is the negated sum of the binary losses divided by the total number of binary learners.

• If Decoding is 'lossweighted', then NegLoss(i,k) is the negated sum of the binary losses divided by the number of binary learners for the kth class.

For more details, see Binary Loss.

PBScore — Positive-class scores

numeric matrix

Positive-class scores for each binary learner, returned as a numeric matrix. PBScore is an n-by-B matrix, where n is the number of observations (size(Mdl.X,1)) and B is the number of binary learners (size(Mdl.CodingMatrix,2)).

Posterior — Posterior class probabilities

numeric matrix

Posterior class probabilities, returned as a numeric matrix. Posterior is an n-by-K matrix, where n is the number of observations (size(Mdl.X,1)) and K is the number of unique classes (size(Mdl.ClassNames,1)).

To request Posterior, you must set 'FitPosterior',true when training the ECOC model using fitcecoc. Otherwise, the software throws an error.

More About

Binary Loss

The binary loss is a function of the class and classification score that determines how well a binary learner classifies an observation into the class. The decoding scheme of an ECOC model specifies how the software aggregates the binary losses and determines the predicted class for each observation.

Assume the following:

• m_kj is element (k,j) of the coding design matrix M—that is, the code corresponding to class k of binary learner j. M is a K-by-B matrix, where K is the number of classes, and B is the number of binary learners.

• s_j is the score of binary learner j for an observation.

• g is the binary loss function.

• $$\widehat{k}$$ is the predicted class for the observation.

The software supports two decoding schemes:

• Loss-based decoding [3] (Decoding is "lossbased") — The predicted class of an observation corresponds to the class that produces the minimum average of the binary losses over all binary learners.

  $$\widehat{k}=\underset{k}{\text{argmin}}\frac{1}{B}{\displaystyle \sum _{j=1}^B\left|m_{kj}\right|g}(m_{kj},s_j).$$

• Loss-weighted decoding [4] (Decoding is "lossweighted") — The predicted class of an observation corresponds to the class that produces the minimum average of the binary losses over the binary learners for the corresponding class.

  $$\widehat{k}=\underset{k}{\text{argmin}}\frac{{\displaystyle \sum _{j=1}^B\left|m_{kj}\right|g}(m_{kj},s_j)}{{\displaystyle \sum }_{j=1}^B\left|m_{kj}\right|}.$$

  The denominator corresponds to the number of binary learners for class k. [1] suggests that loss-weighted decoding improves classification accuracy by keeping loss values for all classes in the same dynamic range.

The predict, resubPredict, and kfoldPredict functions return the negated value of the objective function of argmin as the second output argument (NegLoss) for each observation and class.

This table summarizes the supported binary loss functions, where y_j is a class label for a particular binary learner (in the set {–1,1,0}), s_j is the score for observation j, and g(y_j,s_j) is the binary loss function.

Value	Description	Score Domain	g(yj,sj)
"binodeviance"	Binomial deviance	(–∞,∞)	log[1 + exp(–2yjsj)]/[2log(2)]
"exponential"	Exponential	(–∞,∞)	exp(–yjsj)/2
"hamming"	Hamming	[0,1] or (–∞,∞)	[1 – sign(yjsj)]/2
"hinge"	Hinge	(–∞,∞)	max(0,1 – yjsj)/2
"linear"	Linear	(–∞,∞)	(1 – yjsj)/2
"logit"	Logistic	(–∞,∞)	log[1 + exp(–yjsj)]/[2log(2)]
"quadratic"	Quadratic	[0,1]	[1 – yj(2sj – 1)]2/2

The software normalizes binary losses so that the loss is 0.5 when y_j = 0, and aggregates using the average of the binary learners [1].

Do not confuse the binary loss with the overall classification loss (specified by the LossFun name-value argument of the resubLoss and resubPredict object functions), which measures how well an ECOC classifier performs as a whole.

Algorithms

The software can estimate class posterior probabilities by minimizing the Kullback-Leibler divergence or by using quadratic programming. For the following descriptions of the posterior estimation algorithms, assume that:

• m_kj is the element (k,j) of the coding design matrix M.

• I is the indicator function.

• $${\widehat{p}}_k$$ is the class posterior probability estimate for class k of an observation, k = 1,...,K.

• r_j is the positive-class posterior probability for binary learner j. That is, r_j is the probability that binary learner j classifies an observation into the positive class, given the training data.

Posterior Estimation Using Kullback-Leibler Divergence

By default, the software minimizes the Kullback-Leibler divergence to estimate class posterior probabilities. The Kullback-Leibler divergence between the expected and observed positive-class posterior probabilities is

$$\Delta (r,\widehat{r})={\displaystyle \sum _{j=1}^L{w}_j}\left[r_j\log \frac{r_j}{{\widehat{r}}_j}+\left(1-r_j\right)\log \frac{1-r_j}{1-{\widehat{r}}_j}\right],$$

where $$w_j={\displaystyle \sum _{S_j}{w}_i^{\ast }}$$ is the weight for binary learner j.

• S_j is the set of observation indices on which binary learner j is trained.

• $$w_i^{\ast }$$ is the weight of observation i.

The software minimizes the divergence iteratively. The first step is to choose initial values $${\widehat{p}}_k^{(0)};k=1,\mathrm{...},K$$ for the class posterior probabilities.

• If you do not specify 'NumKLIterations', then the software tries both sets of deterministic initial values described next, and selects the set that minimizes Δ.

  • $${\widehat{p}}_k^{(0)}=1/K;k=1,\mathrm{...},K.$$

  • $${\widehat{p}}_k^{(0)};k=1,\mathrm{...},K$$ is the solution of the system

    $$M_{01}{\widehat{p}}^{(0)}=r,$$

    where M₀₁ is M with all m_kj = –1 replaced with 0, and r is a vector of positive-class posterior probabilities returned by the L binary learners [Dietterich et al.]. The software uses lsqnonneg to solve the system.

• If you specify 'NumKLIterations',c, where c is a natural number, then the software does the following to choose the set $${\widehat{p}}_k^{(0)};k=1,\mathrm{...},K$$, and selects the set that minimizes Δ.

  • The software tries both sets of deterministic initial values as described previously.

  • The software randomly generates c vectors of length K using rand, and then normalizes each vector to sum to 1.

At iteration t, the software completes these steps:

1. Compute

   $${\widehat{r}}_j^{(t)}=\frac{{\displaystyle \sum _{k=1}^K{\widehat{p}}_k^{(t)}}I(m_{kj}=+1)}{{\displaystyle \sum _{k=1}^K{\widehat{p}}_k^{(t)}}I(m_{kj}=+1\cup m_{kj}=-1)}.$$

2. Estimate the next class posterior probability using

   $${\widehat{p}}_k^{(t+1)}={\widehat{p}}_k^{(t)}\frac{{\displaystyle \sum _{j=1}^L{w}_j}\left[r_{j}I\left(m_{kj}=+1\right)+\left(1-r_j\right)I\left(m_{kj}=-1\right)\right]}{{\displaystyle \sum _{j=1}^L{w}_j}\left[{\widehat{r}}_j^{(t)}I\left(m_{kj}=+1\right)+\left(1-{\widehat{r}}_j^{(t)}\right)I\left(m_{kj}=-1\right)\right]}.$$

3. Normalize $${\widehat{p}}_k^{(t+1)};k=1,\mathrm{...},K$$ so that they sum to 1.

4. Check for convergence.

For more details, see [Hastie et al.] and [Zadrozny].

Posterior Estimation Using Quadratic Programming

Posterior probability estimation using quadratic programming requires an Optimization Toolbox license. To estimate posterior probabilities for an observation using this method, the software completes these steps:

1. Estimate the positive-class posterior probabilities, r_j, for binary learners j = 1,...,L.

2. Using the relationship between r_j and $${\widehat{p}}_k$$ [Wu et al.], minimize

   $${\displaystyle \sum }_{j=1}^L{\left[-r_j{\displaystyle \sum _{k=1}^K{\widehat{p}}_k}I\left(m_{kj}=-1\right)+\left(1-r_j\right){\displaystyle \sum _{k=1}^K{\widehat{p}}_k}I\left(m_{kj}=+1\right)\right]}^2$$

   with respect to $${\widehat{p}}_k$$ and the restrictions

   $$\begin{array}{l}0\le {\widehat{p}}_k\le 1\\ {\displaystyle \sum _k{\widehat{p}}_k}=1.\end{array}$$

   The software performs minimization using quadprog (Optimization Toolbox).

References

[1] Allwein, E., R. Schapire, and Y. Singer. “Reducing multiclass to binary: A unifying approach for margin classifiers.” Journal of Machine Learning Research. Vol. 1, 2000, pp. 113–141.

[2] Dietterich, T., and G. Bakiri. “Solving Multiclass Learning Problems Via Error-Correcting Output Codes.” Journal of Artificial Intelligence Research. Vol. 2, 1995, pp. 263–286.

[3] Escalera, S., O. Pujol, and P. Radeva. “Separability of ternary codes for sparse designs of error-correcting output codes.” Pattern Recog. Lett. Vol. 30, Issue 3, 2009, pp. 285–297.

[4] Escalera, S., O. Pujol, and P. Radeva. “On the decoding process in ternary error-correcting output codes.” IEEE Transactions on Pattern Analysis and Machine Intelligence. Vol. 32, Issue 7, 2010, pp. 120–134.

[5] Hastie, T., and R. Tibshirani. “Classification by Pairwise Coupling.” Annals of Statistics. Vol. 26, Issue 2, 1998, pp. 451–471.

[6] Wu, T. F., C. J. Lin, and R. Weng. “Probability Estimates for Multi-Class Classification by Pairwise Coupling.” Journal of Machine Learning Research. Vol. 5, 2004, pp. 975–1005.

[7] Zadrozny, B. “Reducing Multiclass to Binary by Coupling Probability Estimates.” NIPS 2001: Proceedings of Advances in Neural Information Processing Systems 14, 2001, pp. 1041–1048.

Extended Capabilities

Automatic Parallel Support

Accelerate code by automatically running computation in parallel using Parallel Computing Toolbox™.

To run in parallel, specify the Options name-value argument in the call to this function and set the UseParallel field of the options structure to true using statset:

Options=statset(UseParallel=true)

For more information about parallel computing, see Run MATLAB Functions with Automatic Parallel Support (Parallel Computing Toolbox).

GPU Arrays

Accelerate code by running on a graphics processing unit (GPU) using Parallel Computing Toolbox™.

This function fully supports GPU arrays. For more information, see Run MATLAB Functions on a GPU (Parallel Computing Toolbox).

Version History

Introduced in R2014b