SalzbergWeightLearning

java.lang.Object
- Core.BaseLearners.LearningAttributeWeights.LearningAttributeWeights
- - Core.BaseLearners.LearningAttributeWeights.SalzbergWeightLearning

```
public class SalzbergWeightLearning
extends LearningAttributeWeights
```
This class implements the Slzberg (1991) method for learning attributes weight. This method consists of: - If the classification is correct, then + Weight of coincident attributes is increased + Weight of differing attributes is decreased - If the classification is wrong, then + Weight of coincident attributes is decreased + Weight of differing attributes is increased The increase or decrease is inversely proportional to the distance between test and train individuals and proportional to the classification error Note that this methodology uses feedback, therefore it is needed to classify test cases one after the other

Field Summary
- Fields inherited from class Core.BaseLearners.LearningAttributeWeights.LearningAttributeWeights
  path

Constructors
Constructor and Description
`SalzbergWeightLearning()`

Methods
Modifier and Type	Method and Description
`Attribute[]`	`learn(CBR cbr, Attribute[] attributes)` Mètode de aprenentatge de pesos que modifica el pes dels atributs de la Population train en funció de la Population de test
`void`	`setAcceptableDifference(float diff)` Sets the acceptable difference between numerical values in the classification.
`void`	`setDiscPercent(int p)` Sets the percentage (0,100] of numeric values every group of discretization will contain.
`void`	`setDiscRatio(float r)` Sets the ratio (0,1] of numeric values every group of discretization will contain.
`void`	`setIncrement(float inc)` Set Salzberg's algorithm increment

Methods inherited from class Core.BaseLearners.LearningAttributeWeights.LearningAttributeWeights
addIgnoredWeights, learnWeights

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Class SalzbergWeightLearning