rmontanana/Platform
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix predict_proba in AdaBoost

Browse Source
This commit is contained in:
Ricardo Montañana Gómez
2025-06-18 14:18:15 +02:00
parent 56af1a5f85
commit 4e18dc87be
2 changed files with 36 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

47
src/experimental_clfs/AdaBoost.cpp
View File

@@ -74,32 +74,53 @@ namespace bayesnet {
break; // Stop boosting
}
// Calculate alpha (estimator weight) using SAMME formula
// alpha = log((1 - err) / err) + log(K - 1)
double alpha = std::log((1.0 - weighted_error) / weighted_error) +
std::log(static_cast<double>(n_classes - 1));
// Check for perfect classification BEFORE calculating alpha
if (weighted_error <= 1e-10) {
if (debug) std::cout << " Perfect classification achieved (error=" << weighted_error << ")" << std::endl;
// For perfect classification, use a large but finite alpha
double alpha = 10.0 + std::log(static_cast<double>(n_classes - 1));
// Store the estimator and its weight
models.push_back(std::move(estimator));
alphas.push_back(alpha);
// Update sample weights
updateSampleWeights(models.back().get(), alpha);
if (debug) {
std::cout << "Iteration " << iter << ":" << std::endl;
std::cout << " Weighted error: " << weighted_error << std::endl;
std::cout << " Alpha (finite): " << alpha << std::endl;
std::cout << " Random guess error: " << random_guess_error << std::endl;
}
// Normalize weights
break; // Stop training as we have a perfect classifier
}
// Calculate alpha (estimator weight) using SAMME formula
// alpha = log((1 - err) / err) + log(K - 1)
// Clamp weighted_error to avoid division by zero and infinite alpha
double clamped_error = std::max(1e-15, std::min(1.0 - 1e-15, weighted_error));
double alpha = std::log((1.0 - clamped_error) / clamped_error) +
std::log(static_cast<double>(n_classes - 1));
// Clamp alpha to reasonable bounds to avoid numerical issues
alpha = std::max(-10.0, std::min(10.0, alpha));
// Store the estimator and its weight
models.push_back(std::move(estimator));
alphas.push_back(alpha);
// Update sample weights (only if this is not the last iteration)
if (iter < n_estimators - 1) {
updateSampleWeights(models.back().get(), alpha);
normalizeWeights();
}
if (debug) {
std::cout << "Iteration " << iter << ":" << std::endl;
std::cout << " Weighted error: " << weighted_error << std::endl;
std::cout << " Alpha: " << alpha << std::endl;
std::cout << " Random guess error: " << random_guess_error << std::endl;
}
// Check for perfect classification
if (weighted_error < 1e-10) {
if (debug) std::cout << " Perfect classification achieved, stopping" << std::endl;
break;
std::cout << " Random guess error: " << random_guess_error << std::endl;
}
}

5
tests/TestAdaBoost.cpp
View File

@@ -184,10 +184,9 @@ TEST_CASE("AdaBoost Basic Functionality", "[AdaBoost]")
}
// Check that predict_proba matches the expected predict value
// REQUIRE(pred == (p[0] > p[1] ? 0 : 1));
REQUIRE(pred == (p[0] > p[1] ? 0 : 1));
}
double accuracy = static_cast<double>(correct) / n_samples;
std::cout << "Probability accuracy: " << accuracy << std::endl;
REQUIRE(accuracy > 0.99); // Should achieve good accuracy on this simple dataset
}
}
@@ -711,6 +710,8 @@ TEST_CASE("AdaBoost SAMME Algorithm Validation", "[AdaBoost]")
for (size_t i = 0; i < predictions.size(); i++) {
int pred = predictions[i];
auto probs = probabilities[i];
INFO("Sample " << i << ": predicted=" << pred
<< ", probabilities=[" << probs[0] << ", " << probs[1] << "]");
REQUIRE(pred == (probs[0] > probs[1] ? 0 : 1));
REQUIRE(probs[0] + probs[1] == Catch::Approx(1.0).epsilon(1e-6));