Skip to content

MLP

MLP - GetActivationFunction

Returns the activation function.

IN

Name Type Description
InMLP DataModel_MLP The input MLP
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutActivationFunction Enum < ActivationFunction > Activation function.
Error ErrorState Gets the execution error message

MLP - GetClasses

Returns the number of classes considered by the model.

IN

Name Type Description
InMLP DataModel_MLP The input MLP
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutInt Int Number of classes.
Error ErrorState Gets the execution error message

MLP - GetFeatures

Returns the number of features considered by the model.

IN

Name Type Description
InMLP DataModel_MLP The input MLP
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutInt Int Number of features.
Error ErrorState Gets the execution error message

MLP - GetHiddenLayerSizes

Returns the hidden layer sizes.

IN

Name Type Description
InMLP DataModel_MLP The input MLP
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutInt IntArray Hidden layers.
Error ErrorState Gets the execution error message

MLP - IsNormalized

Returns whether to normalize input features.

IN

Name Type Description
InMLP DataModel_MLP The input MLP
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutBool Bool True if input features are normalized, false otherwise.
Error ErrorState Gets the execution error message

MLP - IsTrained

Checks if model has been trained.

IN

Name Type Description
InMLP DataModel_MLP The input MLP
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutBool Bool true if model was trained, false otherwise.
Error ErrorState Gets the execution error message

MLP - Load

Load internal model.

IN

Name Type Description
InFilename PathFile Filename.
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutMLP DataModel_MLP Loaded model.
Error ErrorState Gets the execution error message

MLP - Predict_Dataset

Classify dataset's features.

IN

Name Type Description
InMLP DataModel_MLP The input MLP
InDataset Dataset Input dataset.
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutInt IntArray Predicted classes.
Error ErrorState Gets the execution error message

MLP - Predict_Multiple

Classify multiple feature vectors.

IN

Name Type Description
InMLP DataModel_MLP The input MLP
InFeatures FloatArrayArray Input features.
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutInt IntArray Predicted classes.
Error ErrorState Gets the execution error message

MLP - Predict_Single

Classify a feature vector.

IN

Name Type Description
InMLP DataModel_MLP The input MLP
InFeatures FloatArray Input features.
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutInt Int Predicted class.
Error ErrorState Gets the execution error message

MLP - Save

Save internal model.

IN

Name Type Description
InMLP DataModel_MLP The input MLP
InFilename PathFile Filename.
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
Error ErrorState Gets the execution error message

MLP - Train_FromDataset

Train a new model.

IN

Name Type Description
InDataset Dataset Dataset.
InActivationFunction Enum < ActivationFunction > Activation function.
InNormalizeFeatures Bool Whether to normalize input features.
InIterationCount Int Number of training iterations. A value between 500 and 2000 is suggested.
InLearningRate Float Learning rate (how much the weights are changed at each step). 1e-3 is a good value to start from.
InMomentum Float Value helping the training process by causing the weight updates to continue in the same direction they were already going. It adds a fraction of the previous update to the current update. With 0.1 the new update will be only slightly influenced by the previous one, with 0.9 it will be heavily influenced.
InValidationRatio Float Validation ratio (0 means no validation). Range: 〔0,1).
InShuffle Bool Whether to shuffle the data.
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutMLP DataModel_MLP Trained model.
Error ErrorState Gets the execution error message

MLP - Train_FromFeaturesAndClasses

Train a new model.

IN

Name Type Description
InFeatures FloatArrayArray Features.
InClasses IntArray Corresponding classes.
InNClasses Int Number of classes.
InActivationFunction Enum < ActivationFunction > Activation function.
InNormalizeFeatures Bool Whether to normalize input features.
InIterationCount Int Number of training iterations. A value between 500 and 2000 is suggested.
InLearningRate Float Learning rate (how much the weights are changed at each step). 1e-3 is a good value to start from.
InMomentum Float Value helping the training process by causing the weight updates to continue in the same direction they were already going. It adds a fraction of the previous update to the current update. With 0.1 the new update will be only slightly influenced by the previous one, with 0.9 it will be heavily influenced.
InValidationRatio Float Validation ratio (0 means no validation). Range: 〔0,1).
InShuffle Bool Whether to shuffle the data.
Name String Set the tool's name
Enable Bool Sets if current tool is enabled or not

OUT

Name Type Description
OutMLP DataModel_MLP Trained model.
Error ErrorState Gets the execution error message