WOLFRAM

Wolfram Language & System Documentation Center

IndeterminateThreshold

is an option for Classify, Predict, and related functions that specifies below what probability or probability density a result should be considered indeterminate.

Details
Details and Options Details and Options
Examples  
Basic Examples  
Scope  
Properties & Relations  
See Also
History
Cite this Page

IndeterminateThreshold

IndeterminateThreshold

is an option for Classify, Predict, and related functions that specifies below what probability or probability density a result should be considered indeterminate.

Details

  • With IndeterminateThresholdp0, Indeterminate will be returned if the probability or probability density of the predicted value is less than p0.

Examples

open all close all

Basic Examples  (2)

Train a logistic classifier:

Wolfram Language code: c = Classify[{1 -> "a", 2 -> "b", 3 -> "a", 4 -> "b", 5 -> "b"}, Method -> "LogisticRegression"]

Evaluate the most probable class for a new example:

Wolfram Language code: c[2.5]

Obtain the actual class probabilities:

Wolfram Language code: c[2.5, "Probabilities"]

Set an IndeterminateThreshold to the class probability for the classification:

Wolfram Language code: c[2.5, IndeterminateThreshold -> 0.9]

Train a linear predictor:

Wolfram Language code: p = Predict[{1 -> 0.24, 2 -> 0.31, 3 -> 0.5, 4 -> 0.6}, Method -> "LinearRegression"]

Obtain the probability distribution for a given example:

Wolfram Language code: p[2.5, "Distribution"]

Predict the most probable value:

Wolfram Language code: p[2.5]

Set an IndeterminateThreshold to the probability density for the prediction:

Wolfram Language code: p[2.5, IndeterminateThreshold -> 20]

Scope  (8)

Set the default probability threshold at classifier training:

Wolfram Language code: c = Classify[{1 -> "a", 2 -> "b", 3 -> "a", 4 -> "b", 5 -> "b"}, Method -> "LogisticRegression", IndeterminateThreshold -> 0.9]

Obtain the classification:

Wolfram Language code: c[2.5]

Extract the IndeterminateThreshold from a ClassifierFunction:

Wolfram Language code: Information[ClassifierFunction[Association["ExampleNumber" -> 5, "ClassNumber" -> 2, "Input" -> Association["Preprocessor" -> MachineLearning`MLProcessor["ToMLDataset", Association["Input" -> Association["f1" -> Association["Type" -> "Numerical"]], ... te" -> DateObject[{2026, 1, 22, 17, 43, 46.655315`8.42147610258836}, "Instant", "Gregorian", 1.], "ProcessorCount" -> 4, "ProcessorType" -> "x86-64", "OperatingSystem" -> "MacOSX", "SystemWordLength" -> 64, "Evaluations" -> {}]]], IndeterminateThreshold]

Obtain a ClassifierFunction with a modified IndeterminateThreshold value:

Wolfram Language code: Classify[ClassifierFunction[Association["ExampleNumber" -> 5, "ClassNumber" -> 2, "Input" -> Association["Preprocessor" -> MachineLearning`MLProcessor["ToMLDataset", Association["Input" -> Association["f1" -> Association["Type" -> "Numerical"]], ... te" -> DateObject[{2026, 1, 22, 17, 43, 46.655315`8.42147610258836}, "Instant", "Gregorian", 1.], "ProcessorCount" -> 4, "ProcessorType" -> "x86-64", "OperatingSystem" -> "MacOSX", "SystemWordLength" -> 64, "Evaluations" -> {}]]], IndeterminateThreshold -> 0.3]

Set the IndeterminateThreshold value for a built-in classifier decision:

Wolfram Language code: Classify["FacebookTopic", "I love carrots!"]
Wolfram Language code: Classify["FacebookTopic", "I love carrots!", IndeterminateThreshold -> 0.9]

Set the probability threshold at predictor training:

Wolfram Language code: p = Predict[{1 -> 0.24, 2 -> 0.31, 3 -> 0.5, 4 -> 0.6}, Method -> "LinearRegression", IndeterminateThreshold -> 20]

Obtain the prediction:

Wolfram Language code: p[2.5]

Extract the current IndeterminateThreshold value from a PredictorFunction:

Wolfram Language code: Information[PredictorFunction[Association["ExampleNumber" -> 4, "Input" -> Association["Preprocessor" -> MachineLearning`MLProcessor["ToMLDataset", Association["Input" -> Association["f1" -> Association["Type" -> "Numerical"]], "Output" -> As ... e" -> DateObject[{2026, 1, 22, 17, 9, 11.084835`7.7973042190975175}, "Instant", "Gregorian", 1.], "ProcessorCount" -> 4, "ProcessorType" -> "x86-64", "OperatingSystem" -> "MacOSX", "SystemWordLength" -> 64, "Evaluations" -> {}]]], IndeterminateThreshold]

Obtain a PredictorFunction with a modified IndeterminateThreshold value:

Wolfram Language code: Predict[PredictorFunction[Association["ExampleNumber" -> 4, "Input" -> Association["Preprocessor" -> MachineLearning`MLProcessor["ToMLDataset", Association["Input" -> Association["f1" -> Association["Type" -> "Numerical"]], "Output" -> As ... e" -> DateObject[{2026, 1, 22, 17, 9, 11.084835`7.7973042190975175}, "Instant", "Gregorian", 1.], "ProcessorCount" -> 4, "ProcessorType" -> "x86-64", "OperatingSystem" -> "MacOSX", "SystemWordLength" -> 64, "Evaluations" -> {}]]], IndeterminateThreshold -> 0.3]

Change the value of IndeterminateThreshold in a built-in classifier:

Wolfram Language code: Predict["NameAge"]["John"]
Wolfram Language code: Predict["NameAge", "John", IndeterminateThreshold -> 0.5]

Properties & Relations  (1)

Train a linear predictor:

Wolfram Language code: p = Predict[{1 -> 0.24, 2 -> 0.31, 3 -> 0.5, 4 -> 0.6}, Method -> "LinearRegression"]

Visualize the probability density for a given example:

Wolfram Language code: Plot[PDF[p[2.5, "Distribution"]][x], {x, 0, 1}, PlotRange -> All]

For this example, the predicted value and its corresponding probability density are as follows:

Wolfram Language code: p[2.5]
Wolfram Language code: pdf[p[2.5]]

If the threshold is higher than the probability density, no value is predicted:

Wolfram Language code: p[2.5, IndeterminateThreshold -> 20]
Wolfram Research (2014), IndeterminateThreshold, Wolfram Language function, https://reference.wolfram.com/language/ref/IndeterminateThreshold.html.

Text

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

CMS

Wolfram Language. 2014. "IndeterminateThreshold." Wolfram Language & System Documentation Center. Wolfram Research. https://reference.wolfram.com/language/ref/IndeterminateThreshold.html.

APA

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

BibTeX

@misc{reference.wolfram_2026_indeterminatethreshold, author="Wolfram Research", title="{IndeterminateThreshold}", year="2014", howpublished="\url{https://reference.wolfram.com/language/ref/IndeterminateThreshold.html}", note=[Accessed: 13-June-2026]}

BibLaTeX

@online{reference.wolfram_2026_indeterminatethreshold, organization={Wolfram Research}, title={IndeterminateThreshold}, year={2014}, url={https://reference.wolfram.com/language/ref/IndeterminateThreshold.html}, note=[Accessed: 13-June-2026]}