smoteTabularSynthesizer
Description
To generate synthetic data, you can first create a
smoteTabularSynthesizer object using an existing multivariate data set.
Then, use the synthesizeTabularData object function to synthesize data using SMOTE (synthetic
minority oversampling technique). After you synthesize data, you can test whether the new data
set comes from the same distribution as the original data set. Use the mmdtest or
knntest function to
determine how close the data distributions are to each other. For more information about
SMOTE, see Generate Synthetic Data Using SMOTE.
Creation
Syntax
Description
synthesizer = smoteTabularSynthesizer(X)synthesizer) using the
existing data X.
synthesizer = smoteTabularSynthesizer(___,Name=Value)
Input Arguments
Name-Value Arguments
Properties
Object Functions
synthesizeTabularData | Synthesize tabular data using binning-based or SMOTE-based synthesizer |
Examples
Use existing training data to create a smoteTabularSynthesizer object. Then, synthesize data using the synthesizeTabularData object function. Train a model using the existing training data, and then train the same type of model using both the existing training data and the synthetic data. Compare the performance of the two models using test data.
Load the carbig data set, which contains measurements of cars made in the 1970s and early 1980s. Categorize the cars based on whether they were made in Europe.
load carbig Origin = categorical(cellstr(Origin)); Origin = mergecats(Origin,["France","Germany", ... "Sweden","Italy","England"],"Europe"); Origin = mergecats(Origin,["USA","Japan"],"NotEurope"); tabulate(Origin)
Value Count Percent
Europe 73 17.98%
NotEurope 333 82.02%
The data is imbalanced, with only about 18% of cars originating in Europe.
Create a table containing the variables Acceleration, Displacement, and so on, as well as the response variable Origin. Remove rows of cars where the table has missing values.
cars = table(Acceleration,Displacement,Horsepower, ...
MPG,Weight,Origin);
cars = rmmissing(cars);Partition the data into training and test sets. Use approximately 50% of the observations for model training and synthesizing new data, and 50% of the observations for model testing. Use stratified partitioning so that approximately the same ratio of European to non-European cars exists in both the training and test sets.
rng("default")
cv = cvpartition(cars.Origin,Holdout=0.5);
trainCars = cars(training(cv),:);
testCars = cars(test(cv),:);Create a smoteTabularSynthesizer object using the trainCars data set. Specify Origin as the class labels variable.
synthesizer = smoteTabularSynthesizer(trainCars,"Origin")synthesizer =
smoteTabularSynthesizer
VariableNames: ["Acceleration" "Displacement" "Horsepower" "MPG" "Weight"]
ClassNames: [Europe NotEurope]
NumNeighbors: [5 5]
Distance: "seuclidean"
NumObservations: [34 162]
Properties, Methods
synthesizer is a smoteTabularSynthesizer object with two classes (Europe and NotEurope).
Synthesize new data by using synthesizer. Specify to generate 40 observations belonging to the class of European cars only.
syntheticCars = synthesizeTabularData(synthesizer,40,ClassNames="Europe"); The synthesizeTabularData object function uses the information stored in synthesizer to generate syntheticCars.
To visualize the difference between the existing European car data and the synthetic European car data, you can use the detectdrift function. Filter the trainCars data to include European car data only. The detectdrift function uses permutation testing to detect drift between europeanCars and syntheticCars.
europeanCars = trainCars(trainCars.Origin=="Europe",:);
dd = detectdrift(europeanCars,syntheticCars);dd is a DriftDiagnostics object with a plotEmpiricalCDF object function for visualization.
For the continuous variables, use the plotEmpiricalCDF function to see the difference between the empirical cumulative distribution function (ecdf) of the values in europeanCars and the ecdf of the values in syntheticCars.
continuousVariable ="Acceleration"; plotEmpiricalCDF(dd,Variable=continuousVariable) legend(["Real data","Synthetic data"])
For the Acceleration predictor, the ecdf plot for the existing values (in blue) matches the ecdf plot for the synthetic values (in red) fairly well.
Train an SVM classifier using the original training data trainCars. Specify Origin as the response variable, and standardize the predictors before training. Then, train the same kind of classifier using both the original data and the synthetic data (syntheticCars).
originalMdl = fitcsvm(trainCars,"Origin",Standardize=true); newMdl = fitcsvm([trainCars;syntheticCars],"Origin",Standardize=true);
Evaluate the performance of the two models on the test set using confusion matrices.
originalPredictions = predict(originalMdl,testCars); newPredictions = predict(newMdl,testCars); tiledlayout(1,2) nexttile confusionchart(testCars.Origin,originalPredictions) title("Original Model") nexttile confusionchart(testCars.Origin,newPredictions) title("New Model")
The model trained on the original data classifies all test observations as non-European cars. The model trained on the original and synthetic data has greater accuracy than the other model and correctly classifies the majority of European cars in the test set.
Evaluate data synthesized from an existing data set. Compare the existing and synthetic data sets for each class to determine the similarity between the two multivariate data distributions.
Load the sample file fisheriris.csv, which contains iris data including sepal length, sepal width, petal width, and species type. Read the file into a table, and then display the first eight observations in the table.
fisheriris = readtable("fisheriris.csv");
head(fisheriris) SepalLength SepalWidth PetalLength PetalWidth Species
___________ __________ ___________ __________ __________
5.1 3.5 1.4 0.2 {'setosa'}
4.9 3 1.4 0.2 {'setosa'}
4.7 3.2 1.3 0.2 {'setosa'}
4.6 3.1 1.5 0.2 {'setosa'}
5 3.6 1.4 0.2 {'setosa'}
5.4 3.9 1.7 0.4 {'setosa'}
4.6 3.4 1.4 0.3 {'setosa'}
5 3.4 1.5 0.2 {'setosa'}
Display the types of irises in the table and their proportion in the data set.
tabulate(fisheriris.Species)
Value Count Percent
setosa 50 33.33%
versicolor 50 33.33%
virginica 50 33.33%
The data set contains three types of irises (setosa, versicolor, and virginica) with 50 observations each.
Create separate tables for the versicolor data and the virginica data. Combine the two tables into one.
versicolorData = fisheriris(fisheriris.Species=="versicolor",:); virginicaData = fisheriris(fisheriris.Species=="virginica",:); irisData = [versicolorData;virginicaData]
irisData=100×5 table
SepalLength SepalWidth PetalLength PetalWidth Species
___________ __________ ___________ __________ ______________
7 3.2 4.7 1.4 {'versicolor'}
6.4 3.2 4.5 1.5 {'versicolor'}
6.9 3.1 4.9 1.5 {'versicolor'}
5.5 2.3 4 1.3 {'versicolor'}
6.5 2.8 4.6 1.5 {'versicolor'}
5.7 2.8 4.5 1.3 {'versicolor'}
6.3 3.3 4.7 1.6 {'versicolor'}
4.9 2.4 3.3 1 {'versicolor'}
6.6 2.9 4.6 1.3 {'versicolor'}
5.2 2.7 3.9 1.4 {'versicolor'}
5 2 3.5 1 {'versicolor'}
5.9 3 4.2 1.5 {'versicolor'}
6 2.2 4 1 {'versicolor'}
6.1 2.9 4.7 1.4 {'versicolor'}
5.6 2.9 3.6 1.3 {'versicolor'}
6.7 3.1 4.4 1.4 {'versicolor'}
⋮
Create 200 new observations from the data in irisData: 100 for the versicolor class and 100 for the virginica class. First, create an object by using the smoteTabularSynthesizer function. Then, synthesize the data by using the synthesizeTabularData object function.
To better compare class-specific data later on, call the synthesizeTabularData function twice and return versicolor and virginica synthetic data separately.
synthesizer = smoteTabularSynthesizer(irisData,"Species"); syntheticVersicolorData = synthesizeTabularData(synthesizer,100, ... ClassNames="versicolor"); syntheticVirginicaData = synthesizeTabularData(synthesizer,100, ... ClassNames="virginica");
For each type of iris, visually compare the observations in the existing data set and the synthetic data set by using scatter plots. Each point corresponds to an observation. The point color indicates the species of the corresponding iris (blue for versicolor and red for virginica).
tiledlayout(2,2) nexttile scatter(versicolorData,"SepalLength","PetalLength") title("Existing Versicolor Data") nexttile scatter(syntheticVersicolorData,"SepalLength","PetalLength") title("Synthetic Versicolor Data") nexttile scatter(virginicaData,"SepalLength","PetalLength", ... MarkerEdgeColor="red") title("Existing Virginica Data") nexttile scatter(syntheticVirginicaData,"SepalLength","PetalLength", ... MarkerEdgeColor="red") title("Synthetic Virginica Data")
For each iris type, the scatter plots indicate that the existing data set and the synthetic data set have similar characteristics.
For each iris type, compare the existing and synthetic data sets by using the knntest function. The function performs a two-sample hypothesis test for the null hypothesis that the data sets come from the same distribution.
[knnstat1,p1,h1] = knntest(versicolorData,syntheticVersicolorData)
knnstat1 = 0.5327
p1 = 0.8830
h1 = 0
[knnstat2,p2,h2] = knntest(virginicaData,syntheticVirginicaData)
knnstat2 = 0.5400
p2 = 0.7738
h2 = 0
For each iris type, the returned value of h = 0 indicates that knntest fails to reject the null hypothesis that the existing data set and the synthetic data set come from different distributions at the 5% significance level. As with other hypothesis tests, this result does not guarantee that the null hypothesis is true. That is, the data sets do not necessarily come from the same distribution, but the high p-value indicates that the distributions of the existing and synthetic data sets are similar.
Algorithms
References
[1] Chawla, Nitesh V., Kevin W. Bowyer, Lawrence O. Hall, and W. Philip Kegelmeyer. "SMOTE: Synthetic Minority Over-sampling Technique." Journal of Artificial Intelligence Research 16 (2002): 321-357.
Version History
Introduced in R2026a
