private double trainingStats(EvalParameters evalParams)
{
int numCorrect = 0;
for (int ei = 0; ei < numUniqueEvents; ei++)
{
for (int ni = 0; ni < this.numTimesEventsSeen[ei]; ni++)
{
double[] modelDistribution = new double[numOutcomes];
if (values != null)
{
PerceptronModel.eval(contexts[ei], values[ei], modelDistribution, evalParams, false);
}
else
{
PerceptronModel.eval(contexts[ei], null, modelDistribution, evalParams, false);
}
int max = maxIndex(modelDistribution);
if (max == outcomeList[ei])
{
numCorrect++;
}
}
}
double trainingAccuracy = (double)numCorrect / numEvents;
display("Stats: (" + numCorrect + "/" + numEvents + ") " + trainingAccuracy + "\n");
return(trainingAccuracy);
}
public virtual void nextIteration(int iteration)
{
iteration--; //move to 0-based index
int numCorrect = 0;
int oei = 0;
int si = 0;
IDictionary <string, float?>[] featureCounts = (IDictionary <string, float?>[]) new IDictionary[numOutcomes];
for (int oi = 0; oi < numOutcomes; oi++)
{
featureCounts[oi] = new Dictionary <string, float?>();
}
PerceptronModel model = new PerceptronModel(parameters, predLabels, pmap, outcomeLabels);
foreach (Sequence <Event> sequence in sequenceStream)
{
Event[] taggerEvents = sequenceStream.updateContext(sequence, model);
Event[] events = sequence.Events;
bool update = false;
for (int ei = 0; ei < events.Length; ei++, oei++)
{
if (!taggerEvents[ei].Outcome.Equals(events[ei].Outcome))
{
update = true;
//break;
}
else
{
numCorrect++;
}
}
if (update)
{
for (int oi = 0; oi < numOutcomes; oi++)
{
featureCounts[oi].Clear();
}
//System.err.print("train:");for (int ei=0;ei<events.length;ei++) {System.err.print(" "+events[ei].getOutcome());} System.err.println();
//training feature count computation
for (int ei = 0; ei < events.Length; ei++, oei++)
{
string[] contextStrings = events[ei].Context;
float[] values = events[ei].Values;
int oi = (int)omap[events[ei].Outcome];
for (int ci = 0; ci < contextStrings.Length; ci++)
{
float value = 1;
if (values != null)
{
value = values[ci];
}
float?c = featureCounts[oi][contextStrings[ci]];
if (c == null)
{
c = value;
}
else
{
c += value;
}
featureCounts[oi][contextStrings[ci]] = c;
}
}
//evaluation feature count computation
//System.err.print("test: ");for (int ei=0;ei<taggerEvents.length;ei++) {System.err.print(" "+taggerEvents[ei].getOutcome());} System.err.println();
foreach (Event taggerEvent in taggerEvents)
{
string[] contextStrings = taggerEvent.Context;
float[] values = taggerEvent.Values;
int oi = (int)omap[taggerEvent.Outcome];
for (int ci = 0; ci < contextStrings.Length; ci++)
{
float value = 1;
if (values != null)
{
value = values[ci];
}
float?c = featureCounts[oi][contextStrings[ci]];
if (c == null)
{
c = -1 * value;
}
else
{
c -= value;
}
if (c == 0f)
{
featureCounts[oi].Remove(contextStrings[ci]);
}
else
{
featureCounts[oi][contextStrings[ci]] = c;
}
}
}
for (int oi = 0; oi < numOutcomes; oi++)
{
foreach (string feature in featureCounts[oi].Keys)
{
int pi = pmap.get(feature);
if (pi != -1)
{
//System.err.println(si+" "+outcomeLabels[oi]+" "+feature+" "+featureCounts[oi].get(feature));
parameters[pi].updateParameter(oi, (double)featureCounts[oi][feature]);
if (useAverage)
{
if (updates[pi][oi][VALUE] != 0)
{
averageParams[pi].updateParameter(oi,
updates[pi][oi][VALUE] *
(numSequences * (iteration - updates[pi][oi][ITER]) +
(si - updates[pi][oi][EVENT])));
//System.err.println("p avp["+pi+"]."+oi+"="+averageParams[pi].getParameters()[oi]);
}
//System.err.println("p updates["+pi+"]["+oi+"]=("+updates[pi][oi][ITER]+","+updates[pi][oi][EVENT]+","+updates[pi][oi][VALUE]+") + ("+iteration+","+oei+","+params[pi].getParameters()[oi]+") -> "+averageParams[pi].getParameters()[oi]);
updates[pi][oi][VALUE] = (int)parameters[pi].Parameters[oi];
updates[pi][oi][ITER] = iteration;
updates[pi][oi][EVENT] = si;
}
}
}
}
model = new PerceptronModel(parameters, predLabels, pmap, outcomeLabels);
}
si++;
}
//finish average computation
double totIterations = (double)iterations * si;
if (useAverage && iteration == iterations - 1)
{
for (int pi = 0; pi < numPreds; pi++)
{
double[] predParams = averageParams[pi].Parameters;
for (int oi = 0; oi < numOutcomes; oi++)
{
if (updates[pi][oi][VALUE] != 0)
{
predParams[oi] += updates[pi][oi][VALUE] *
(numSequences * (iterations - updates[pi][oi][ITER]) -
updates[pi][oi][EVENT]);
}
if (predParams[oi] != 0)
{
predParams[oi] /= totIterations;
averageParams[pi].setParameter(oi, predParams[oi]);
//System.err.println("updates["+pi+"]["+oi+"]=("+updates[pi][oi][ITER]+","+updates[pi][oi][EVENT]+","+updates[pi][oi][VALUE]+") + ("+iterations+","+0+","+params[pi].getParameters()[oi]+") -> "+averageParams[pi].getParameters()[oi]);
}
}
}
}
display(". (" + numCorrect + "/" + numEvents + ") " + ((double)numCorrect / numEvents) + "\n");
}
private MutableContext[] findParameters(int iterations, bool useAverage)
{
display("Performing " + iterations + " iterations.\n");
int[] allOutcomesPattern = new int[numOutcomes];
for (int oi = 0; oi < numOutcomes; oi++)
{
allOutcomesPattern[oi] = oi;
}
/// <summary>
/// Stores the estimated parameter value of each predicate during iteration. </summary>
MutableContext[] parameters = new MutableContext[numPreds];
for (int pi = 0; pi < numPreds; pi++)
{
parameters[pi] = new MutableContext(allOutcomesPattern, new double[numOutcomes]);
for (int aoi = 0; aoi < numOutcomes; aoi++)
{
parameters[pi].setParameter(aoi, 0.0);
}
}
EvalParameters evalParams = new EvalParameters(parameters, numOutcomes);
/// <summary>
/// Stores the sum of parameter values of each predicate over many iterations. </summary>
MutableContext[] summedParams = new MutableContext[numPreds];
if (useAverage)
{
for (int pi = 0; pi < numPreds; pi++)
{
summedParams[pi] = new MutableContext(allOutcomesPattern, new double[numOutcomes]);
for (int aoi = 0; aoi < numOutcomes; aoi++)
{
summedParams[pi].setParameter(aoi, 0.0);
}
}
}
// Keep track of the previous three accuracies. The difference of
// the mean of these and the current training set accuracy is used
// with tolerance to decide whether to stop.
double prevAccuracy1 = 0.0;
double prevAccuracy2 = 0.0;
double prevAccuracy3 = 0.0;
// A counter for the denominator for averaging.
int numTimesSummed = 0;
double stepsize = 1;
for (int i = 1; i <= iterations; i++)
{
// Decrease the stepsize by a small amount.
if (stepSizeDecrease != null)
{
stepsize *= 1 - stepSizeDecrease.GetValueOrDefault();
}
displayIteration(i);
int numCorrect = 0;
for (int ei = 0; ei < numUniqueEvents; ei++)
{
int targetOutcome = outcomeList[ei];
for (int ni = 0; ni < this.numTimesEventsSeen[ei]; ni++)
{
// Compute the model's prediction according to the current parameters.
double[] modelDistribution = new double[numOutcomes];
if (values != null)
{
PerceptronModel.eval(contexts[ei], values[ei], modelDistribution, evalParams, false);
}
else
{
PerceptronModel.eval(contexts[ei], null, modelDistribution, evalParams, false);
}
int maxOutcome = maxIndex(modelDistribution);
// If the predicted outcome is different from the target
// outcome, do the standard update: boost the parameters
// associated with the target and reduce those associated
// with the incorrect predicted outcome.
if (maxOutcome != targetOutcome)
{
for (int ci = 0; ci < contexts[ei].Length; ci++)
{
int pi = contexts[ei][ci];
if (values == null)
{
parameters[pi].updateParameter(targetOutcome, stepsize);
parameters[pi].updateParameter(maxOutcome, -stepsize);
}
else
{
parameters[pi].updateParameter(targetOutcome, stepsize * values[ei][ci]);
parameters[pi].updateParameter(maxOutcome, -stepsize * values[ei][ci]);
}
}
}
// Update the counts for accuracy.
if (maxOutcome == targetOutcome)
{
numCorrect++;
}
}
}
// Calculate the training accuracy and display.
double trainingAccuracy = (double)numCorrect / numEvents;
if (i < 10 || (i % 10) == 0)
{
display(". (" + numCorrect + "/" + numEvents + ") " + trainingAccuracy + "\n");
}
// TODO: Make averaging configurable !!!
bool doAveraging;
if (useAverage && useSkippedlAveraging && (i < 20 || isPerfectSquare(i)))
{
doAveraging = true;
}
else if (useAverage)
{
doAveraging = true;
}
else
{
doAveraging = false;
}
if (doAveraging)
{
numTimesSummed++;
for (int pi = 0; pi < numPreds; pi++)
{
for (int aoi = 0; aoi < numOutcomes; aoi++)
{
summedParams[pi].updateParameter(aoi, parameters[pi].Parameters[aoi]);
}
}
}
// If the tolerance is greater than the difference between the
// current training accuracy and all of the previous three
// training accuracies, stop training.
if (Math.Abs(prevAccuracy1 - trainingAccuracy) < tolerance &&
Math.Abs(prevAccuracy2 - trainingAccuracy) < tolerance &&
Math.Abs(prevAccuracy3 - trainingAccuracy) < tolerance)
{
display("Stopping: change in training set accuracy less than " + tolerance + "\n");
break;
}
// Update the previous training accuracies.
prevAccuracy1 = prevAccuracy2;
prevAccuracy2 = prevAccuracy3;
prevAccuracy3 = trainingAccuracy;
}
// Output the final training stats.
trainingStats(evalParams);
// Create averaged parameters
if (useAverage)
{
for (int pi = 0; pi < numPreds; pi++)
{
for (int aoi = 0; aoi < numOutcomes; aoi++)
{
summedParams[pi].setParameter(aoi, summedParams[pi].Parameters[aoi] / numTimesSummed);
}
}
return(summedParams);
}
else
{
return(parameters);
}
}