lcanals.tc.core
Class LinearRegressionEstimator

java.lang.Object
  |
  +--lcanals.tc.core.LinearRegressionEstimator
All Implemented Interfaces:
Estimator, java.io.Serializable

public class LinearRegressionEstimator
extends java.lang.Object
implements Estimator, java.io.Serializable

Author:
www.lfcanals.es.org
See Also:
Serialized Form

Inner Class Summary
 class LinearRegressionEstimator.Information
          Class for internal state access
 
Constructor Summary
LinearRegressionEstimator(float performance, int size, float k)
          Builds a new estimator
 
Method Summary
 Estimation estimateOnlineTask(PlanTask task)
          The same as estimateTask but updating user data if task is finished
 Estimation estimateTask(Task task)
          estimatedProjectTime = f(d) where f(x)=a+bx maximumEstimatedProjectTime = (1+ r/4)·f(d)
 EstimatorInformation getInternalInformation()
          Returns a specific object with internal information about the estimator
 java.lang.String getType()
          Says the name of this type of Estimator
 void learnTask(PlanTask planTask)
          To learn a new element, the oldest one must be forgotten.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

LinearRegressionEstimator

public LinearRegressionEstimator(float performance,
                                 int size,
                                 float k)
Builds a new estimator
Parameters:
size - number of amples to be considered (memory size)
Method Detail

estimateOnlineTask

public Estimation estimateOnlineTask(PlanTask task)
The same as estimateTask but updating user data if task is finished
Specified by:
estimateOnlineTask in interface Estimator

estimateTask

public Estimation estimateTask(Task task)
estimatedProjectTime = f(d) where f(x)=a+bx maximumEstimatedProjectTime = (1+ r/4)·f(d)
Specified by:
estimateTask in interface Estimator

learnTask

public void learnTask(PlanTask planTask)
To learn a new element, the oldest one must be forgotten. Learning a new element implies to recalculate a and b using regression.
Specified by:
learnTask in interface Estimator

getType

public java.lang.String getType()
Description copied from interface: Estimator
Says the name of this type of Estimator
Specified by:
getType in interface Estimator

getInternalInformation

public EstimatorInformation getInternalInformation()
Description copied from interface: Estimator
Returns a specific object with internal information about the estimator
Specified by:
getInternalInformation in interface Estimator


Luis F. Canals. Supported by www.sourceforge.net