WOLFRAM

Wolfram Language & System Documentation Center
Wolfram Language Home Page »

ClassifierFunction
✖
ClassifierFunction

✖

represents a function generated by Classify that classifies data into classes.

Details and Options

  • ClassifierFunction works like Function.
  • ClassifierFunction[…][data] attempts to classify data, returning the class in which data is considered most likely to be.
  • ClassifierFunction[…][{data1,data2,…}] attempts to classify all the datai.
  • ClassifierFunction[…][data,prop] gives the specified property of the classification associated with data.
  • Possible properties applicable to all methods include:
  • "Decision"best class according to the probabilities and the utility function
    "TopProbabilities"probabilities for most likely classes
    "TopProbabilities"nprobabilities for the n most likely classes
    "Probability"classprobability for a specific class
    "Probabilities"association of probabilities for all possible classes
    "SHAPValues"Shapley additive feature explanations for each example
    "Properties"list of all properties available
  • "SHAPValues" assesses the contribution of features by comparing predictions with different sets of features removed and then synthesized. The option MissingValueSynthesis can be used to specify how the missing features are synthesized. SHAP explanations are given as odds ratio multipliers with respect to the class training prior. "SHAPValues"n can be used to control the number of samples used for the numeric estimations of SHAP explanations.
  • ClassifierFunction[…][data,…,opts] specifies that the classifier should use the options opts when applied to data.
  • Possible options are:
  • ClassPriors Automaticexplicit prior probabilities for classes
    IndeterminateThreshold Automaticbelow what probability to return Indeterminate
    MissingValueSynthesisAutomatichow to synthesize missing values
    PerformanceGoalAutomaticwhich aspect of performance to optimize
    TargetDevice "CPU"the target device on which to perform training
    RecalibrationFunction Automatichow to post-process class probabilities
    UtilityFunction Automaticutility expressed as a function of actual and predicted class
  • A ClassifierFunction[…] trained in an older version of the Wolfram Language will still work in the current version.
  • Classify[net] can be used to convert a NetChain or NetGraph representing a classifier into a ClassifierFunction[…].
  • Classify[ClassifierFunction[…],opts], can be used to update the values of PerformanceGoal, ClassPriors, IndeterminateThreshold, UtilityFunction or FeatureExtractor of the classifier.
  • In Classify[ClassifierFunction[…],FeatureExtractorfe], the FeatureExtractorFunction[…] fe will be prepended to the existing feature extractor.
  • Information[ClassifierFunction[…]] generates an information panel about the classifier and its estimated performances.
  • Information[ClassifierFunction[…],prop] can be used to obtain specific properties.
  • Information of a ClassifierFunction may include the following properties:
  • "Accuracy"estimated accuracy of the classifier
    "BatchEvaluationTime"marginal time to predict one example when a batch is given
    "Classes"list of classes that the classifier can return
    "ClassNumber"number of classes that the classifier can return
    "EvaluationTime"time needed to classify one example
    "ExampleNumber"number of training examples
    "FeatureTypes"feature types of the classfier input
    "FunctionMemory"memory needed to store the classifier
    "FunctionProperties"all classification properties available for this classifier
    "IndeterminateThreshold"value of IndeterminateThreshold used by the classifier
    "LearningCurve"performance as a function of the training set size
    "MaxTrainingMemory"maximum memory used during training
    "MeanCrossEntropy"estimated mean cross entropy of the classifier
    "Method"value of Method used by the classifier
    "MethodDescription"summary of the method
    "MethodOption"full method option to be reused in a new training
    "MethodParameters"parameter settings of the method
    "Properties"all information properties available for this classifier
    "FeatureExtractor"feature extractor as FeatureExtractorFunction
    "TrainingClassPriors"the distribution of classes seen during training
    "TrainingTime"time used by Classify to generate the classifier
    "UtilityFunction"value of UtilityFunction used by the classifier
  • Information properties also include all method suboptions.

Examples

open allclose all

Basic Examples  (2)Summary of the most common use cases

Create a ClassifierFunction with Classify and a list of labeled examples:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-b9dzea
Out[1]=1

Classify an unlabeled example with the ClassifierFunction:

In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-l0fphr
Out[2]=2

Classify multiple examples:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-i8ukky
Out[3]=3

Return the probabilities of the classes given the feature of an example:

In[4]:=4
✖
https://wolfram.com/xid/0g7e5i7yiky-5t4ktv
Out[4]=4

Return the sorted probabilities of the most likely classes:

In[5]:=5
✖
https://wolfram.com/xid/0g7e5i7yiky-klyxkz
Out[5]=5

Return the probability of the most probable class:

In[6]:=6
✖
https://wolfram.com/xid/0g7e5i7yiky-o7xoyn
Out[6]=6

Return the probability of a given class:

In[7]:=7
✖
https://wolfram.com/xid/0g7e5i7yiky-bgq0f
Out[7]=7

Plot the probability of class "B" as a function of the feature:

In[8]:=8
✖
https://wolfram.com/xid/0g7e5i7yiky-oa6egz
Out[8]=8

Generate a ClassifierFunction using multiple features:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-6d3z0t
Out[1]=1

Use the function on a new example:

In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-q87iwn
Out[2]=2

Classify an example that has missing features:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-pjaio8
Out[3]=3

Get the probabilities for the most probable classes:

In[4]:=4
✖
https://wolfram.com/xid/0g7e5i7yiky-v3ik3x
Out[4]=4

Scope  (5)Survey of the scope of standard use cases

Create a function classifying textual data:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-vcqdub
Out[1]=1

Classify new examples:

In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-hhhr2x
Out[2]=2

Obtain information on the function:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-jn5tqy
Out[3]=3

Obtain the properties that can be used by this function:

In[4]:=4
✖
https://wolfram.com/xid/0g7e5i7yiky-kwj0qq
Out[4]=4

Train a classifier function:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-vvh6kg
Out[1]=1

Generate a classifier measurements object of the function applied to a test set:

In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-duh5q9
Out[2]=2

Get the accuracy from the function on the test set:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-cbhd3s
Out[3]=3

Visualize the confusion matrix:

In[4]:=4
✖
https://wolfram.com/xid/0g7e5i7yiky-4pfth2
Out[4]=4

Generate a classifier function whose input is an association:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-pl8yra
Out[1]=1

Use the function on an example:

In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-10r04a
Out[2]=2

Classify examples containing missing features:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-deouc6
Out[3]=3

Train a classifier:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-ooqlup
Out[1]=1

Store the ClassifierFunction[…] into a file using the "WMLF" format:

In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-qro8f2
In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-ezpjam
In[4]:=4
✖
https://wolfram.com/xid/0g7e5i7yiky-9goiy9
Out[4]=4

Load the classifier from the file using Import:

In[5]:=5
✖
https://wolfram.com/xid/0g7e5i7yiky-iuml56
Out[5]=5

Use the loaded classifier on new data:

In[6]:=6
✖
https://wolfram.com/xid/0g7e5i7yiky-u2wn4p
Out[6]=6
In[7]:=7
✖
https://wolfram.com/xid/0g7e5i7yiky-8dxsp7
Out[7]=7

Train a classifier to predict a person's odds of surviving or dying in the Titanic crash:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-t1tf33
Out[1]=1

Calculate the prior odds of a passenger dying:

In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-of8sjz
Out[2]=2

Use the classifier to predict the odds of a person dying:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-0z7ebi
Out[3]=3

Get an explanation of how each feature multiplied the model's predicted odds of a class:

In[4]:=4
✖
https://wolfram.com/xid/0g7e5i7yiky-h0rxx8
Out[4]=4

Compare the model's explanation of feature impact to the base rate odds:

In[5]:=5
✖
https://wolfram.com/xid/0g7e5i7yiky-b1m3d3
Out[5]=5
Out[5]=5

Options  (6)Common values & functionality for each option

ClassPriors  (1)

Train a classifier on an imbalanced dataset:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-g33b9h
In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-8p5txa
Out[2]=2

The example 5False is classified as True:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-s227oj
Out[3]=3
In[4]:=4
✖
https://wolfram.com/xid/0g7e5i7yiky-ia8b5e
Out[4]=4

Classify this example with a uniform prior over classes:

In[5]:=5
✖
https://wolfram.com/xid/0g7e5i7yiky-kafd4s
Out[5]=5
In[6]:=6
✖
https://wolfram.com/xid/0g7e5i7yiky-bj1v9f
Out[6]=6

The class priors of a classifier can also be updated after training:

In[7]:=7
✖
https://wolfram.com/xid/0g7e5i7yiky-45vl1r
Out[7]=7
In[8]:=8
✖
https://wolfram.com/xid/0g7e5i7yiky-9fyrxe
Out[8]=8

IndeterminateThreshold  (1)

Train a classifier:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-f26y4p
In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-2l2lc2
Out[2]=2

Obtain class probabilities for an example:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-lfqvzb
Out[3]=3

The most probable class is chosen as the prediction:

In[4]:=4
✖
https://wolfram.com/xid/0g7e5i7yiky-u7mhh
Out[4]=4

No prediction is made if no class probabilities exceed a specified probability threshold:

In[5]:=5
✖
https://wolfram.com/xid/0g7e5i7yiky-wrck2
Out[5]=5

Update the value of the threshold permanently:

In[6]:=6
✖
https://wolfram.com/xid/0g7e5i7yiky-m6emr6
Out[6]=6
In[7]:=7
✖
https://wolfram.com/xid/0g7e5i7yiky-r6tarl
Out[7]=7

RecalibrationFunction  (2)

Train a classifier function:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-4aizph
Out[1]=1

Compute the class probabilities of a new example:

In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-yagshs
Out[2]=2

Check if the model has been calibrated:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-nw6el1
Out[3]=3

Temporarily set a recalibration function to apply to the probabilities:

In[4]:=4
✖
https://wolfram.com/xid/0g7e5i7yiky-nh9ww
Out[4]=4

Set a permanent recalibration function to apply to the probabilities:

In[5]:=5
✖
https://wolfram.com/xid/0g7e5i7yiky-yghi42
Out[5]=5

Compute the class probabilities of a new example:

In[6]:=6
✖
https://wolfram.com/xid/0g7e5i7yiky-p9pr10
Out[6]=6

Remove the recalibration function from the classifier:

✖
https://wolfram.com/xid/0g7e5i7yiky-yyc7ct

Load the Titanic dataset:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-2pli6l

Create a nearest neighbors classifier with no calibration function:

In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-dcnt2l
Out[2]=2

The classifier is slightly overconfident:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-fxmvjl
Out[3]=3

Select the worst classification case in the test set:

In[4]:=4
✖
https://wolfram.com/xid/0g7e5i7yiky-87l5rk
Out[4]=4

Evaluate the estimated probabilities:

In[5]:=5
✖
https://wolfram.com/xid/0g7e5i7yiky-0gt5cu
Out[5]=5

Use "temperature scaling" to reduce the classifier self-confidence:

In[6]:=6
✖
https://wolfram.com/xid/0g7e5i7yiky-6tt1id
Out[6]=6
In[7]:=7
✖
https://wolfram.com/xid/0g7e5i7yiky-x8okg7
Out[7]=7

TargetDevice  (1)

Train a classifier using a neural network:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-gnlysk

Evaluate the resulting classifier on system's default GPU and look at its AbsoluteTiming:

In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-4fgmey

Compare the previous timing with the one achieved by using the default CPU computation:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-g2ypu8

UtilityFunction  (1)

Train a classifier:

In[1]:=1
✖
https://wolfram.com/xid/0g7e5i7yiky-h3wasx
In[2]:=2
✖
https://wolfram.com/xid/0g7e5i7yiky-lowgr7
Out[2]=2

By default, the most probable class is predicted:

In[3]:=3
✖
https://wolfram.com/xid/0g7e5i7yiky-gwhw1
Out[3]=3
In[4]:=4
✖
https://wolfram.com/xid/0g7e5i7yiky-omckhm
Out[4]=4

Specify a utility function that penalizes examples of class "yes" being misclassified as "no":

In[5]:=5
✖
https://wolfram.com/xid/0g7e5i7yiky-lqwpd
Out[5]=5

Update the value of the utility function permanently:

In[6]:=6
✖
https://wolfram.com/xid/0g7e5i7yiky-ho2veh
Out[6]=6
In[7]:=7
✖
https://wolfram.com/xid/0g7e5i7yiky-clqi9j
Out[7]=7
Wolfram Research (2014), ClassifierFunction, Wolfram Language function, https://reference.wolfram.com/language/ref/ClassifierFunction.html (updated 2021).
✖
Wolfram Research (2014), ClassifierFunction, Wolfram Language function, https://reference.wolfram.com/language/ref/ClassifierFunction.html (updated 2021).

Text

Wolfram Research (2014), ClassifierFunction, Wolfram Language function, https://reference.wolfram.com/language/ref/ClassifierFunction.html (updated 2021).

✖
Wolfram Research (2014), ClassifierFunction, Wolfram Language function, https://reference.wolfram.com/language/ref/ClassifierFunction.html (updated 2021).

CMS

Wolfram Language. 2014. "ClassifierFunction." Wolfram Language & System Documentation Center. Wolfram Research. Last Modified 2021. https://reference.wolfram.com/language/ref/ClassifierFunction.html.

✖
Wolfram Language. 2014. "ClassifierFunction." Wolfram Language & System Documentation Center. Wolfram Research. Last Modified 2021. https://reference.wolfram.com/language/ref/ClassifierFunction.html.

APA

Wolfram Language. (2014). ClassifierFunction. Wolfram Language & System Documentation Center. Retrieved from https://reference.wolfram.com/language/ref/ClassifierFunction.html

✖
Wolfram Language. (2014). ClassifierFunction. Wolfram Language & System Documentation Center. Retrieved from https://reference.wolfram.com/language/ref/ClassifierFunction.html

BibTeX

@misc{reference.wolfram_2025_classifierfunction, author="Wolfram Research", title="{ClassifierFunction}", year="2021", howpublished="\url{https://reference.wolfram.com/language/ref/ClassifierFunction.html}", note=[Accessed: 25-April-2025 ]}

✖
@misc{reference.wolfram_2025_classifierfunction, author="Wolfram Research", title="{ClassifierFunction}", year="2021", howpublished="\url{https://reference.wolfram.com/language/ref/ClassifierFunction.html}", note=[Accessed: 25-April-2025 ]}

BibLaTeX

@online{reference.wolfram_2025_classifierfunction, organization={Wolfram Research}, title={ClassifierFunction}, year={2021}, url={https://reference.wolfram.com/language/ref/ClassifierFunction.html}, note=[Accessed: 25-April-2025 ]}

✖
@online{reference.wolfram_2025_classifierfunction, organization={Wolfram Research}, title={ClassifierFunction}, year={2021}, url={https://reference.wolfram.com/language/ref/ClassifierFunction.html}, note=[Accessed: 25-April-2025 ]}