Modifier and Type | Method and Description |
---|---|
double |
getContinuous()
Gets the amount continuous distributions are adjusted during learning.
|
double |
getDiscrete()
Gets the amount distributions containing discrete variables are adjusted during learning.
|
DiscretePriorMethod |
getDiscretePriorMethod()
The default discrete prior to use for discrete distributions during parameter learning.
|
boolean |
getIncludeGlobalCovariance()
When Gaussian distributions are adjusted according to the
getContinuous() prior, this property determines whether the global covariance should be included in the adjustment, as well as the global variance. |
double |
getSimpleVariance()
Used to make a fixed adjustment to all covariance matrices during learning, by increasing each diagonal (variance) entry.
|
void |
setContinuous(double value)
Sets the amount continuous distributions are adjusted during learning.
|
void |
setDiscrete(double value)
Sets the amount distributions containing discrete variables are adjusted during learning.
|
void |
setDiscretePriorMethod(DiscretePriorMethod value)
The default discrete prior to use for discrete distributions during parameter learning.
|
void |
setIncludeGlobalCovariance(boolean value)
When Gaussian distributions are adjusted according to the
getContinuous() prior, this property determines whether the global covariance should be included in the adjustment, as well as the global variance. |
void |
setSimpleVariance(double value)
Used to make a fixed adjustment to all covariance matrices during learning, by increasing each diagonal (variance) entry.
|
String |
toString()
Returns a
String that represents this instance. |
void |
zeroAll()
Sets all values to zero.
|
public void zeroAll()
public double getSimpleVariance()
public void setSimpleVariance(double value)
public DiscretePriorMethod getDiscretePriorMethod()
public void setDiscretePriorMethod(DiscretePriorMethod value)
public double getContinuous()
This value is used to avoid boundary conditions, such as perfect correlations.
The larger the number of cases used during learning, the less impact this value has.
The value defines the number of virtual cases taken from the global statistics (overall data summary statistics), that are included when learning continuous Gaussian distributions. The property getIncludeGlobalCovariance()
determines how the adjustments are made.
Setting this value to zero, will disable the adjustments.
public void setContinuous(double value)
This value is used to avoid boundary conditions, such as perfect correlations.
The larger the number of cases used during learning, the less impact this value has.
The value defines the number of virtual cases taken from the global statistics (overall data summary statistics), that are included when learning continuous Gaussian distributions. The property getIncludeGlobalCovariance()
determines how the adjustments are made.
Setting this value to zero, will disable the adjustments.
public boolean getIncludeGlobalCovariance()
getContinuous()
prior, this property determines whether the global covariance should be included in the adjustment, as well as the global variance.public void setIncludeGlobalCovariance(boolean value)
getContinuous()
prior, this property determines whether the global covariance should be included in the adjustment, as well as the global variance.public double getDiscrete()
This value is used to avoid boundary conditions.
The larger the number of cases used during learning, the less impact this value has.
The value defines the number of virtual cases taken from the global statistics (overall data summary statistics), that are included when learning distributions with discrete variables.
Setting this value to zero, will disable the adjustments.
public void setDiscrete(double value)
This value is used to avoid boundary conditions.
The larger the number of cases used during learning, the less impact this value has.
The value defines the number of virtual cases taken from the global statistics (overall data summary statistics), that are included when learning distributions with discrete variables.
Setting this value to zero, will disable the adjustments.
Copyright © 2021. All rights reserved.