nuget
Class containing the routines to perform class membership prediction using a trained SVM.
Prediction
01 Syntax
Microsoft.VisualBasic.MachineLearning.SVM.Prediction
02 Methods
| Name | Overloads | Summary |
|---|---|---|
| Predict | 2 | Predict the class for a single input vector. |
| PredictLabels | 1 | Predict the labels for all the data points in a problem. |
| PredictLabelsProbability | 1 | Predict the probability distributions over all labels for each data point in a problem. |
| PredictProbability | 1 | Predicts a class distribution for the single input vector. |
03 Members
Predicts the class memberships of all the vectors in the problem.
Parameters
| Name | Type | Description |
|---|---|---|
problem | Problem | The SVM Problem to solve |
outputFile | String | File for result output |
model | Model | The Model to use |
predict_probability | Boolean | Whether to output a distribution over the classes |
Returns
Percentage correctly labelled
Predict the class for a single input vector.
Parameters
| Name | Type | Description |
|---|---|---|
model | Model | The Model to use for prediction |
x | Node() | The vector for which to predict class |
Returns
The result
Predict the labels for all the data points in a problem.
Parameters
Returns
The predicted labels
Predict the probability distributions over all labels for each data point in a problem.
Parameters
Returns
A distribution over labels for each data point
Predicts a class distribution for the single input vector.
Parameters
| Name | Type | Description |
|---|---|---|
model | Model | Model to use for prediction |
x | Node() | The vector for which to predict the class distribution |
Returns
A probability distribtion over classes