def generalStatistics(trainingData, testData, protectedIndex, protectedValue): print('Size of training data:', len(trainingData)) print('Size of testing data:', len(testData)) print("SP of training data: %f" % statisticalParity(trainingData, protectedIndex, protectedValue)) print("SP of test data: %f" % statisticalParity(testData, protectedIndex, protectedValue))
def statistics(trainingData, testData, learner, protectedIndex, protectedValue): h = learner(trainingData) trainingError = labelError(trainingData, h) testError = labelError(testData, h) sp = statisticalParity(testData, protectedIndex, protectedValue, h) UBIF = individualFairness(trainingData, learner, FLIP_PROPORTION) return trainingError, testError, sp, UBIF
def runBaseline(trainingData, testData, learner, protectedIndex, protectedValue): h = learner(trainingData) print("Training error: %f" % labelError(trainingData, h)) print("Test error: %f" % labelError(testData, h)) print("SP of the hypothesis: %f" % statisticalParity(testData, protectedIndex, protectedValue, h)) UBIF = individualFairness(trainingData, learner, FLIP_PROPORTION) print("UBIF of the hypothesis on training: %f" % UBIF)
def statistics(trainingData, testData, learner, protectedIndex, protectedValue): h = learner(trainingData) trainingError = labelError(trainingData, h) testError = labelError(testData, h) sp = statisticalParity(testData, protectedIndex, protectedValue, h) UBIF = individualFairness(trainingData, learner, FLIP_PROPORTION) return trainingError, testError, sp, UBIF
def makeErrorFunction(protectedIndex, protectedValue, spWeight): ''' Run boosting on a decision stump finder that uses an error function which is a linear combination of statistical imparity and label error w * statisticalParity + (1-w) * labelError ''' sp = lambda data, h: ef.statisticalParity(data, protectedIndex, protectedValue, h=h) le = ef.minLabelErrorOfHypothesisAndNegation return ef.makeLinearCombination(sp, le, spWeight)
def runBaseline(trainingData, testData, learner, protectedIndex, protectedValue): h = learner(trainingData) print("Training error: %f" % labelError(trainingData, h)) print("Test error: %f" % labelError(testData, h)) print("SP of the hypothesis: %f" % statisticalParity(testData, protectedIndex, protectedValue, h)) UBIF = individualFairness(trainingData, learner, FLIP_PROPORTION) print("UBIF of the hypothesis on training: %f" % UBIF)
def makeErrorFunction(protectedIndex, protectedValue, spWeight): sp = lambda data, h: ef.statisticalParity(data, protectedIndex, protectedValue, h=h) le = ef.minLabelErrorOfHypothesisAndNegation return ef.makeLinearCombination(sp, le, spWeight)
def makeErrorFunction(protectedIndex, protectedValue, spWeight): sp = lambda data, h: ef.statisticalParity( data, protectedIndex, protectedValue, h=h) le = ef.minLabelErrorOfHypothesisAndNegation return ef.makeLinearCombination(sp, le, spWeight)
def generalStatistics(trainingData, testData, protectedIndex, protectedValue): print('Size of training data:', len(trainingData)) print('Size of testing data:', len(testData)) print("SP of training data: %f" % statisticalParity(trainingData, protectedIndex, protectedValue)) print("SP of test data: %f" % statisticalParity(testData, protectedIndex, protectedValue))