@Override
public double empiricalLoss(
final org.drip.function.definition.R1ToR1 funcLearnerR1ToR1,
final org.drip.spaces.instance.GeneralizedValidatedVector gvviX,
final org.drip.spaces.instance.GeneralizedValidatedVector gvviY)
throws java.lang.Exception {
if (null == funcLearnerR1ToR1
|| null == gvviX
|| !(gvviX instanceof org.drip.spaces.instance.ValidatedR1)
|| null == gvviY
|| !(gvviY instanceof org.drip.spaces.instance.ValidatedR1))
throw new java.lang.Exception("LpLossLearner::empiricalLoss => Invalid Inputs");
double[] adblX = ((org.drip.spaces.instance.ValidatedR1) gvviX).instance();
double[] adblY = ((org.drip.spaces.instance.ValidatedR1) gvviY).instance();
double dblEmpiricalLoss = 0.;
int iNumSample = adblX.length;
if (iNumSample != adblY.length)
throw new java.lang.Exception("LpLossLearner::empiricalLoss => Invalid Inputs");
for (int i = 0; i < iNumSample; ++i)
dblEmpiricalLoss +=
java.lang.Math.pow(
java.lang.Math.abs(funcLearnerR1ToR1.evaluate(adblX[i]) - adblY[i]),
_dblLossExponent);
return dblEmpiricalLoss / _dblLossExponent;
}
@Override
public double empiricalRisk(
final org.drip.measure.continuous.R1R1 distR1R1,
final org.drip.function.definition.R1ToR1 funcLearnerR1ToR1,
final org.drip.spaces.instance.GeneralizedValidatedVector gvviX,
final org.drip.spaces.instance.GeneralizedValidatedVector gvviY)
throws java.lang.Exception {
if (null == distR1R1
|| null == funcLearnerR1ToR1
|| null == gvviX
|| !(gvviX instanceof org.drip.spaces.instance.ValidatedR1)
|| null == gvviY
|| !(gvviY instanceof org.drip.spaces.instance.ValidatedR1))
throw new java.lang.Exception("LpLossLearner::empiricalRisk => Invalid Inputs");
double[] adblX = ((org.drip.spaces.instance.ValidatedR1) gvviX).instance();
double[] adblY = ((org.drip.spaces.instance.ValidatedR1) gvviY).instance();
double dblNormalizer = 0.;
double dblEmpiricalLoss = 0.;
int iNumSample = adblX.length;
if (iNumSample != adblY.length)
throw new java.lang.Exception("LpLossLearner::empiricalRisk => Invalid Inputs");
for (int i = 0; i < iNumSample; ++i) {
double dblDensity = distR1R1.density(adblX[i], adblY[i]);
dblNormalizer += dblDensity;
dblEmpiricalLoss +=
dblDensity
* java.lang.Math.pow(
java.lang.Math.abs(funcLearnerR1ToR1.evaluate(adblX[i]) - adblY[i]),
_dblLossExponent);
}
return dblEmpiricalLoss / _dblLossExponent / dblNormalizer;
}