|
|||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | ||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |
java.lang.Objectinfer.Infer
tmpLearn.Learner
tmpLearn.NaiveGDLearner
public class NaiveGDLearner
Learner instance using gradient descent.
Note, this class is for debugging and JUNIT test only. For
real circumstance usage, please turn to NaiveDNLearner
.
If you are really a big fan of gradient descent, please
rewrite loadingTrainingData(MCSAT)
according
to that in NaiveDNLearner.loadingTrainingData(MCSAT)
.
Field Summary |
---|
Fields inherited from class tmpLearn.Learner |
---|
backtrackCount_, backtracked, currentD, currentWeight, finalWeight, isFirstTime, isHardMappings, odds, oldWeight, trainingSatisification, trainingViolation |
Fields inherited from class infer.Infer |
---|
db, grounding, mln, options |
Constructor Summary | |
---|---|
NaiveGDLearner()
|
Method Summary | |
---|---|
void |
loadingTrainingData(MCSAT _mcsat)
NEVER USE THIS IN REAL CIRCUMSTANCE... |
boolean |
updateWeight(MCSAT mcsat)
Updating Learner.currentWeight using Gradient
Descent method. |
Methods inherited from class tmpLearn.Learner |
---|
dumpAnswers, fillInCurrentWeight, run, validatingLearningPerformance |
Methods inherited from class infer.Infer |
---|
cleanUp, ground, loadMLN, setUp |
Methods inherited from class java.lang.Object |
---|
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
Constructor Detail |
---|
public NaiveGDLearner()
Method Detail |
---|
public void loadingTrainingData(MCSAT _mcsat)
loadingTrainingData
in class Learner
_mcsat
- The MCSAT object to be filled in.public boolean updateWeight(MCSAT mcsat)
Learner.currentWeight
using Gradient
Descent method.
updateWeight
in class Learner
mcsat
- MCSAT instance after this iteration.
|
|||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | ||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |