Visitar URL original
FeatureExtractorFunction: Apply a feature extraction pipeline—Wolfram Documentation

WOLFRAM

Wolfram Language & System Documentation Center

FeatureExtractorFunction[…]

represents a feature extractor function generated by FeatureExtraction.

Details and Options
Details and Options Details and Options
Examples  
Basic Examples  
Scope  
Possible Issues  
See Also
Related Guides
History
Cite this Page

FeatureExtractorFunction

FeatureExtractorFunction[…]

represents a feature extractor function generated by FeatureExtraction.

Details and Options

  • FeatureExtractorFunction works like Function.
  • FeatureExtractorFunction[…][data] extracts features from data.
  • FeatureExtractorFunction[…][{data1,data2,…}] extracts features from each of the datai.
  • FeatureExtractorFunction[…][data,prop] gives the specified property of the feature extraction associated with data.
  • Possible properties include:
  • "ExtractedFeatures"features extracted from data (default)
    "OriginalData"deduce original data from extracted features
    "ReconstructedData"extraction and inverse extraction of data
  • The following options can be given:
  • PerformanceGoalAutomaticaspects of performance to try to optimize
    RandomSeeding1234what seeding of pseudorandom generators should be done internally
  • Possible settings for RandomSeeding include:
  • Automaticautomatically reseed every time the function is called
    Inheriteduse externally seeded random numbers
    seeduse an explicit integer or strings as a seed

Examples

open all close all

Basic Examples  (1)

Train a FeatureExtractorFunction on a simple dataset:

Wolfram Language code: fe = FeatureExtraction[{{1.4, "A"}, {1.5, "A"}, {2.3, "B"}, {5.4, "B"}}]

Extract features from a new example:

Wolfram Language code: fe[{2.4, "A"}]

Extract features from a list of examples:

Wolfram Language code: fe[{{2.4, "A"}, {3.7, "B"}}]

Scope  (5)

Some feature extractors can only perform an approximation of the inverse extraction:

Wolfram Language code: fe = FeatureExtraction[{{1.4, 1.4, 5.4, 5.2}, {1.5, 1.5, 6.4, 5.2}, {1.2, 1.2, 6.2, 5.2}, {1.6, 1.6, 4.3, 5.2}}, "DimensionReducedVector"]
Wolfram Language code: features = fe[{1.1, 1.2, 5.1, 5.7}]
Wolfram Language code: fe[features, "OriginalData"]

The FeatureExtraction property "ReconstructedData" can be used to obtain the data after extraction and reconstruction:

Wolfram Language code: FeatureExtraction[{{1.4, 1.4, 5.4, 5.2}, {1.5, 1.5, 6.4, 5.2}, {1.2, 1.2, 6.2, 5.2}, {1.6, 1.6, 4.3, 5.2}}, "DimensionReducedVector", "ReconstructedData"]

Some feature extractors cannot be inverted:

Wolfram Language code: FeatureExtraction[{[image], [image] , [image], [image]}, "ImageFeatures", "ReconstructedData"]

Train a feature extractor from a dataset that contains missing values:

Wolfram Language code: fe = FeatureExtraction[{{1.4, Missing[], "A"}, {1.5, 50.2, "A"}, {Missing[], 42.3, "B"}, {5.4, 61.7, "B"}}]

The feature extractor now indicates that missing values are imputed. The feature extractor can extract features even when values are missing:

Wolfram Language code: fe[{3.4, Missing[], "B"}]

Get Information from a trained FeatureExtractorFunction:

Wolfram Language code: Information[FeatureExtractorFunction[Association["ExampleNumber" -> 4, "Preprocessor" -> MachineLearning`MLProcessor["ToMLDataset", Association["Input" -> Association["age" -> Association["Type" -> "Numerical"], "gender" -> Association["Type" - ... ate" -> DateObject[{2025, 5, 2, 13, 0, 39.042795`8.344115878952366}, "Instant", "Gregorian", 2.], "ProcessorCount" -> 10, "ProcessorType" -> "ARM64", "OperatingSystem" -> "MacOSX", "SystemWordLength" -> 64, "Evaluations" -> {}]]]]

Find the available properties:

Wolfram Language code: Information[FeatureExtractorFunction[Association["ExampleNumber" -> 4, "Preprocessor" -> MachineLearning`MLProcessor["ToMLDataset", Association["Input" -> Association["age" -> Association["Type" -> "Numerical"], "gender" -> Association["Type" - ... ate" -> DateObject[{2025, 5, 2, 13, 0, 39.042795`8.344115878952366}, "Instant", "Gregorian", 2.], "ProcessorCount" -> 10, "ProcessorType" -> "ARM64", "OperatingSystem" -> "MacOSX", "SystemWordLength" -> 64, "Evaluations" -> {}]]], "Properties"]

Get information about the feature names and training examples:

Wolfram Language code: Information[FeatureExtractorFunction[Association["ExampleNumber" -> 4, "Preprocessor" -> MachineLearning`MLProcessor["ToMLDataset", Association["Input" -> Association["age" -> Association["Type" -> "Numerical"], "gender" -> Association["Type" - ... ate" -> DateObject[{2025, 5, 2, 13, 0, 39.042795`8.344115878952366}, "Instant", "Gregorian", 2.], "ProcessorCount" -> 10, "ProcessorType" -> "ARM64", "OperatingSystem" -> "MacOSX", "SystemWordLength" -> 64, "Evaluations" -> {}]]], {"FeatureNames", "ExampleNumber"}]

Possible Issues  (1)

Retrieve a old FeatureExtractorFunction using keyed features:

Wolfram Language code: fe = FeatureExtractorFunction[Association["ExampleNumber" -> 0, "Preprocessor" -> MachineLearning`MLProcessor["ToMLDataset", Association["Input" -> Association["input" -> Association["Type" -> "Text"]], "Output" -> Association["f1" -> Asso ... "Date" -> DateObject[{2026, 1, 28, 15, 36, 22.701151`8.10862286062899}, "Instant", "Gregorian", 0.], "ProcessorCount" -> 10, "ProcessorType" -> "ARM64", "OperatingSystem" -> "MacOSX", "SystemWordLength" -> 64, "Evaluations" -> {}]]];

Using absent or incorrect feature names will give errors or missing content:

Wolfram Language code: fe["Hello"]
Wolfram Language code: fe[<|"Input" -> "Hello"|>]

Retrieve the feature names using Information:

Wolfram Language code: Information[fe, "FeatureNames"]

Correct the feature names:

Wolfram Language code: fe[<|"input" -> "Hello"|>]
Wolfram Research (2016), FeatureExtractorFunction, Wolfram Language function, https://reference.wolfram.com/language/ref/FeatureExtractorFunction.html (updated 2017).

Text

Wolfram Research (2016), FeatureExtractorFunction, Wolfram Language function, https://reference.wolfram.com/language/ref/FeatureExtractorFunction.html (updated 2017).

CMS

Wolfram Language. 2016. "FeatureExtractorFunction." Wolfram Language & System Documentation Center. Wolfram Research. Last Modified 2017. https://reference.wolfram.com/language/ref/FeatureExtractorFunction.html.

APA

Wolfram Language. (2016). FeatureExtractorFunction. Wolfram Language & System Documentation Center. Retrieved from https://reference.wolfram.com/language/ref/FeatureExtractorFunction.html

BibTeX

@misc{reference.wolfram_2026_featureextractorfunction, author="Wolfram Research", title="{FeatureExtractorFunction}", year="2017", howpublished="\url{https://reference.wolfram.com/language/ref/FeatureExtractorFunction.html}", note=[Accessed: 14-November-2026]}

BibLaTeX

@online{reference.wolfram_2026_featureextractorfunction, organization={Wolfram Research}, title={FeatureExtractorFunction}, year={2017}, url={https://reference.wolfram.com/language/ref/FeatureExtractorFunction.html}, note=[Accessed: 14-November-2026]}