PredictionEngine
E700880
PredictionEngine is an ML.NET API component that provides a simple, strongly typed interface for making single-record predictions with trained machine learning models in .NET applications.
Statements (47)
| Predicate | Object |
|---|---|
| instanceOf |
ML.NET API component
ⓘ
single-record prediction API ⓘ |
| belongsToEcosystem | .NET machine learning ⓘ |
| createdBy | MLContext NERFINISHED ⓘ |
| createdWithMethod | MLContext.Model.CreatePredictionEngine ⓘ |
| definedInNamespace | Microsoft.ML NERFINISHED ⓘ |
| designGoal |
ease of use for developers new to ML.NET
ⓘ
strong typing of input and output schemas ⓘ |
| discouragedUsage | high-scale web services ⓘ |
| documentationUrl | https://learn.microsoft.com/dotnet/machine-learning/how-to-guides/machine-learning-model-predictions-ml-net ⓘ |
| exampleUsage |
var predEngine = mlContext.Model.CreatePredictionEngine<TData, TPrediction>(model);
ⓘ
var prediction = predEngine.Predict(input); ⓘ |
| hasGenericTypeParameter |
TData
ⓘ
TPrediction ⓘ |
| hasLimitation |
higher allocation cost per prediction
ⓘ
one input at a time ⓘ |
| hasMethod | Predict ⓘ |
| inputType | TData ⓘ |
| lifecycle | short-lived ⓘ |
| operatesOn |
DataView schema
ⓘ
ITransformer models ⓘ |
| outputType | TPrediction ⓘ |
| partOf | ML.NET NERFINISHED ⓘ |
| programmingLanguage | .NET NERFINISHED ⓘ |
| provides |
strongly typed prediction interface
ⓘ
synchronous prediction API ⓘ |
| recommendedUsage |
low-throughput prediction scenarios
ⓘ
single-threaded scenarios ⓘ |
| replacedBy |
ITransformer with batch prediction APIs
ⓘ
PredictionEnginePool ⓘ |
| requires |
schema compatibility between TData and model input
ⓘ
schema compatibility between TPrediction and model output ⓘ trained ML.NET model ⓘ |
| supports |
binary classification predictions
ⓘ
clustering predictions ⓘ multiclass classification predictions ⓘ ranking predictions ⓘ recommendation predictions ⓘ regression predictions ⓘ time series predictions ⓘ |
| supportsDeployment |
console applications
ⓘ
desktop applications ⓘ microservices ⓘ web applications ⓘ |
| threadSafety | not thread-safe ⓘ |
| usedFor |
consuming trained machine learning models
ⓘ
making single-record predictions ⓘ |
Referenced by (1)
Full triples — surface form annotated when it differs from this entity's canonical label.