mbcmodel.modelproperties

Properties and methods for viewing and editing model properties

expand all in page

    Description

    Use these properties and object functions to create and edit model properties.

    Creation

    Syntax

    ModelPropObj = ModelObj.Properties
    properties(ModelObj.Properties)
    ModelObj.Properties.PropertyName = NewValue

    Description

    ModelPropObj = ModelObj.Properties creates an mbcmodel.modelproperties object for ModelObj.

    example

    properties(ModelObj.Properties) lists the property names, types, and allowed values.

    example

    ModelObj.Properties.PropertyName = NewValue changes the ModelObj property name PropertyName to the new value.

    example

    Properties

    expand all

    Every model object mbcmodel.model has an mbcmodel.modelproperties object. In the mbcmodel.modelproperties object, each model type has specific properties.

    These tables provide the properties available for each model type.

    Gaussian Process Models

    PropertyDescriptionValues

    KernelFunction

    Kernel function

    String containing one of these enumerated values:

    • 'Exponential'

    • 'ARDExponential'

    • 'SquaredExponential'

    • 'ARDSquaredExponential'

    • 'Matern32'

    • 'ARDMatern32'

    • 'Matern52'

    • 'ARDMatern52'

    • 'RationalQuadratic'

    • 'ARDRationalQuadratic'

    BasisFunction

    		Explicit basis function

    String containing one of these enumerated values:

    • 'None'

    • 'Constant'

    • 'Linear'

    • 'PureQuadratic'

    Threshold

    Threshold to switch to large data fitting algorithm

    Integer greater than or equal to 1

    ActiveSetSize

    Active set size

    Integer greater than or equal to 1

    ActiveSetMethod

    Large scale active set method

    String containing one of these enumerated values:

    • 'SGMA'

    • 'Entropy'

    • 'Likelihood'

    • 'Random'

    FitMethod

    		Large data fit method

    String containing one of these enumerated values:

    • 'SD'

    • 'FIC'

    • 'SR'

    PredictMethod

    Large data predict method

    String containing one of these enumerated values:

    • 'Exact'

    • 'BCD'

    • 'SD'

    • 'FIC'

    • 'SR'

    InitializeMethod

    Initialize hyperparameters

    String containing one of these enumerated values:

    • 'LOO-loss'

    • 'logML'

    • 'none'

    BoxCox

    Box-Cox transform (power)

    Numeric number between -3 and 3, inclusive

    Linear Models – Polynomial

    PropertyDescriptionValues

    Order

    Polynomial order

    n-by-1 vector of integers greater than or equal to 0, where n is the number of inputs

    InteractionOrder

    Maximum order of interaction terms

    Integer greater than or equal to 0

    TransformInputRange

    Transform inputs

    Boolean

    ParameterNames

    List of parameter names (read-only)

    NA

    StepwiseStatus

    Stepwise status

    Cell containing one of these values:

    • 'Always'

    • 'Never'

    • 'Step'

    BoxCox

    Box-Cox transform (power)

    Numeric number between -3 and 3, inclusive

    Linear Models – Hybrid Spline

    PropertyDescriptionValues

    Order

    Spline and polynomial order vector

    n-by-1 vector of integers greater than or equal to 0 and less than or equal to 3, where n is the number of inputs

    SplineVariable

    Spline variable

    NA

    SplineInteraction

    Order of interaction between spline and polynomial

    Integer greater than or equal to 0 and less than or equal to 3

    Knots

    Position of knots

    Real vector

    ParameterNames

    List of parameter names (read-only)

    NA

    StepwiseStatus

    Stepwise status

    Cell containing one of these values:

    • 'Always'

    • 'Never'

    • 'Step'

    BoxCox

    Box-Cox transform (power)

    Numeric number between -3 and 3, inclusive

    Linear Models – RBF

    PropertyDescriptionValues

    Kernel

    RBF kernel

    String containing one of these enumerated values:

    • 'multiquadric'

    • 'recmultiquadric'

    • 'gaussian'

    • 'thinplate'

    • 'logisticrbf'

    • 'wendland'

    • 'linearrbf'

    • 'cubicrbf'

    Continuity

    Continuity for Wendland kernel

    Integer greater than or equal to 0 and less than or equal to 6

    ParameterNames

    List of parameter names (read-only)

    NA

    StepwiseStatus

    Stepwise status

    Cell containing one of these values:

    • 'Always'

    • 'Never'

    • 'Step'

    BoxCox

    Box-Cox transform (power)

    Numeric number between -3 and 3, inclusive

    Linear Models – Polynomial RBF

    PropertyDescriptionValues

    Order

    		Polynomial order vector

    n-by-1 vector of integers greater than or equal to 0, where n is the number of inputs

    InteractionOrder

    Maximum order of interaction terms

    Integer greater than or equal to 0

    Kernel

    RBF kernel

    String containing one of these enumerated values:

    • 'multiquadric'

    • 'recmultiquadric'

    • 'gaussian'

    • 'thinplate'

    • 'logisticrbf'

    • 'wendland'

    • 'linearrbf'

    • 'cubicrbf'

    Continuity

    Continuity for Wendland kernel

    Integer greater than or equal to 0 and less than or equal to 6

    ParameterNames

    List of parameter names (read-only)

    NA

    StepwiseStatus

    Stepwise status

    Cell containing one of these values:

    • 'Always'

    • 'Never'

    • 'Step'

    BoxCox

    Box-Cox transform (power)

    Numeric number between -3 and 3, inclusive

    Linear Models – Hybrid Spline-RBF

    PropertyDescriptionValues

    Order

    		Spline and polynomial order vector

    n-by-1 vector of integers greater than or equal to 0 and less than or equal to 3, where n is the number of inputs

    SplineVariable

    Spline variable

    NA

    SplineInteraction

    Order of interaction between spline and polynomial

    Integer greater than or equal to 0 and less than or equal to 3

    Knots

    Position of knots

    Real vector

    Kernel

    RBF kernel

    String containing one of these enumerated values:

    • 'multiquadric'

    • 'recmultiquadric'

    • 'gaussian'

    • 'thinplate'

    • 'logisticrbf'

    • 'wendland'

    • 'linearrbf'

    • 'cubicrbf'

    Continuity

    Continuity for Wendland kernel

    Integer greater than or equal to 0 and less than or equal to 6

    ParameterNames

    List of parameter names (read-only)

    NA

    StepwiseStatus

    Stepwise status

    Cell containing one of these values:

    • 'Always'

    • 'Never'

    • 'Step'

    BoxCox

    Box-Cox transform (power)

    Numeric number between -3 and 3, inclusive

    Nonlinear Models – Free Knot Spline

    PropertyDescriptionValues

    Order

    		Spline and polynomial order vector

    n-by-1 vector of integers greater than or equal to 0 and less than or equal to 3, where n is the number of inputs

    NumKnots

    Number of knots

    Positive integer

    Nonlinear Models – Neural Network

    PropertyDescriptionValues

    HiddenLayers

    Number of hidden layers

    Integer value 1 or 2

    Neurons

    Number of Neurons in each hidden layer

    Vector of positive integers

    This property is read-only.

    Type of model objects to be returned to mbcmodel.project, specified as a vector. model.Type returns the model type.

    Note

    Spaces and case in model Type are ignored.

    The model type must be one in this table.

    TypeModel Object
    Polynomialmbcmodel.linearmodel
    Hybrid Splinembcmodel.linearmodel
    RBFmbcmodel.linearmodel
    Hybrid RBFmbcmodel.linearmodel
    Polynomial-RBFmbcmodel.linearmodel
    Hybrid Spline-RBFmbcmodel.linearmodel
    Gaussian Processmbcmodel.model
    Free Knot Splinembcmodel.model
    Transientmbcmodel.model
    User-Definedmbcmodel.model
    Neural Networkmbcmodel.model
    Interpolating RBFmbcmodel.model
    Local Polynomial Splinembcmodel.localmodel
    Local Polynomial with Datummbcmodel.localmodel
    Local Polynomialmbcmodel.localmodel
    Local Hybrid Splinembcmodel.localmodel
    Local Truncated Power Seriesmbcmodel.localmodel
    Local Free Knot Splinembcmodel.localmodel
    Local Multiple Modelsmbcmodel.localmodel
    Local Growth mbcmodel.localmodel
    Local User-Definedmbcmodel.localmodel
    Local Transientmbcmodel.localmodel
    Local Average Fitmbcmodel.localmodel

    You can get a list of types by using getAlternativeTypes. Use this syntax.

    Mlist = getAlternativeTypes(M)

    In this syntax, M is an mbcmodel.model object.

    Data Types: char | string

    This property is read-only.

    Number of inputs to model, specified as a real positive scalar.

    Data Types: double | single

    Examples

    collapse all

    Obtain a mbcmodel.modelproperties object from a model.

    >> ModelObj = mbcmodel.CreateModel('Polynomial', 4);
>> ModelProp = ModelObj.Properties

ModelProp = 

Polynomial Properties
                   Type: 'Polynomial'
              NumInputs: 4
                  Order: [3 3 3 3]
       InteractionOrder: 3
    TransformInputRange: 1
         ParameterNames: {35×1 cell}
         StepwiseStatus: {35×1 cell}
                 BoxCox: 1

    Display the model object.

    >> disp(ModelObj)
   1 + 2*X1 + 17*X2 + 27*X3 + 33*X4 + 3*X1^2 + 8*X1*X2 + 12*X1*X3 + 15*X1*X4 +
   18*X2^2 + 22*X2*X3 + 25*X2*X4 + 28*X3^2 + 31*X3*X4 + 34*X4^2 + 4*X1^3 +    
   5*X1^2*X2 + 6*X1^2*X3 + 7*X1^2*X4 + 9*X1*X2^2 + 10*X1*X2*X3 + 11*X1*X2*X4 +
   13*X1*X3^2 + 14*X1*X3*X4 + 16*X1*X4^2 + 19*X2^3 + 20*X2^2*X3 + 21*X2^2*X4 +
   23*X2*X3^2 + 24*X2*X3*X4 + 26*X2*X4^2 + 29*X3^3 + 30*X3^2*X4 + 32*X3*X4^2 +
   35*X4^3                                                                    
   InputData: [0x4 double]
   OutputData: [0x1 double]
   Status: Not fitted
   Linked to Response: <not linked>

    Obtain the model object properties.

    >> ModelObj = mbcmodel.CreateModel('Polynomial', 4);
>> ModelProp = properties(ModelObj)

ModelProp =

  13×1 cell array

    {'Type'         }
    {'Status'       }
    {'NumInputs'    }
    {'InputData'    }
    {'OutputData'   }
    {'Properties'   }
    {'FitAlgorithm' }
    {'Inputs'       }
    {'Name'         }
    {'Units'        }
    {'Parameters'   }
    {'Response'     }
    {'IsBeingEdited'}

    Create a model and list the model properties.

    >> ModelObj = mbcmodel.CreateModel('RBF',2)
 
ModelObj = 

   A radial basis function network using a multiquadric kernel
   with 0 centers and a global width of 2.                                                  
   The regularization parameter, lambda, is 0.0001.                          
   InputData: [0x2 double]
   OutputData: [0x1 double]
   Status: Not fitted
   Linked to Response: <not linked>

>> properties(ModelObj.Properties)
RBF Properties
              Type: 'RBF'
         NumInputs: 2
            Kernel: 'multiquadric'
        Continuity: 4
    ParameterNames: {0×1 cell}
    StepwiseStatus: {0×1 cell}
            BoxCox: 1

     Kernel: RBF kernel (enum: {'multiquadric','recmultiquadric','gaussian','thinplate',...
          'logisticrbf','wendland','linearrbf','cubicrbf'})
     Continuity: Continuity for Wendland kernel (0,2,4,6) (int: [0,6])
     ParameterNames: List of parameter names (read-only)
     StepwiseStatus: Stepwise status {'Always','Never','Step'} (cell)
     BoxCox: Box-Cox transform (power) (numeric: [-3,3])

    Return the properties as a cell array.

    >> f=ModelObj.Properties.properties

f =

  5×1 cell array

    {'Kernel'        }
    {'Continuity'    }
    {'ParameterNames'}
    {'StepwiseStatus'}
    {'BoxCox'        }

    Version History

    Introduced before R2006a

    See Also

    | | |