Examples with LimeConfig org.kie.kogito.explainability.local.lime.LimeConfig used on opensource projects

Example 61 with LimeConfig

use of org.kie.kogito.explainability.local.lime.LimeConfig in project kogito-apps by kiegroup.

the class ValidationUtilsTest method testStableEval.

@Test
void testStableEval() throws ExecutionException, InterruptedException, TimeoutException, ValidationUtils.ValidationException {
    for (int n = 0; n < 10; n++) {
        Random random = new Random();
        PerturbationContext perturbationContext = new PerturbationContext(4L, random, 1);
        LimeConfig config = new LimeConfig().withPerturbationContext(perturbationContext);
        LimeExplainer explainer = new LimeExplainer(config);
        PredictionProvider model = TestUtils.getSumThresholdModel(0.1, 0.1);
        List<Feature> features = new ArrayList<>();
        for (int i = 0; i < 4; i++) {
            features.add(FeatureFactory.newNumericalFeature("f-" + i, Type.NUMBER.randomValue(perturbationContext).asNumber()));
        }
        PredictionInput input = new PredictionInput(features);
        List<PredictionOutput> outputs = model.predictAsync(List.of(input)).get(Config.DEFAULT_ASYNC_TIMEOUT, Config.DEFAULT_ASYNC_TIMEUNIT);
        Prediction prediction = new SimplePrediction(input, outputs.get(0));
        int topK = 1;
        double posScore = 0.6;
        double minScore = 0.6;
        ValidationUtils.validateLocalSaliencyStability(model, prediction, explainer, topK, posScore, minScore);
    }
}

Example 62 with LimeConfig

use of org.kie.kogito.explainability.local.lime.LimeConfig in project kogito-apps by kiegroup.

the class FraudScoringDmnLimeExplainerTest method testExplanationWithDataDistribution.

@Test
void testExplanationWithDataDistribution() throws ExecutionException, InterruptedException, TimeoutException {
    PredictionProvider model = getModel();
    List<PredictionInput> samples = DmnTestUtils.randomFraudScoringInputs();
    List<PredictionInput> inputs = samples.subList(0, 10);
    Random random = new Random();
    random.setSeed(0);
    PerturbationContext perturbationContext = new PerturbationContext(random, 1);
    LimeConfig initialConfig = new LimeConfig().withSamples(10).withDataDistribution(new PredictionInputsDataDistribution(inputs)).withPerturbationContext(perturbationContext);
    LimeExplainer limeExplainer = new LimeExplainer(initialConfig);
    PredictionInput testPredictionInput = getTestInput();
    List<PredictionOutput> testPredictionOutputs = model.predictAsync(List.of(testPredictionInput)).get(Config.INSTANCE.getAsyncTimeout(), Config.INSTANCE.getAsyncTimeUnit());
    Prediction instance = new SimplePrediction(testPredictionInput, testPredictionOutputs.get(0));
    assertDoesNotThrow(() -> ValidationUtils.validateLocalSaliencyStability(model, instance, limeExplainer, 1, 0.5, 0.5));
}

Example 63 with LimeConfig

use of org.kie.kogito.explainability.local.lime.LimeConfig in project kogito-apps by kiegroup.

the class LoanEligibilityDmnLimeExplainerTest method testExplanationStabilityWithOptimization.

@Test
void testExplanationStabilityWithOptimization() throws ExecutionException, InterruptedException, TimeoutException {
    PredictionProvider model = getModel();
    List<PredictionInput> samples = DmnTestUtils.randomLoanEligibilityInputs();
    List<PredictionOutput> predictionOutputs = model.predictAsync(samples.subList(0, 5)).get();
    List<Prediction> predictions = DataUtils.getPredictions(samples, predictionOutputs);
    long seed = 0;
    LimeConfigOptimizer limeConfigOptimizer = new LimeConfigOptimizer().withDeterministicExecution(true).withStepCountLimit(20);
    Random random = new Random();
    PerturbationContext perturbationContext = new PerturbationContext(seed, random, 1);
    LimeConfig initialConfig = new LimeConfig().withPerturbationContext(perturbationContext);
    LimeConfig optimizedConfig = limeConfigOptimizer.optimize(initialConfig, predictions, model);
    assertThat(optimizedConfig).isNotSameAs(initialConfig);
    LimeExplainer limeExplainer = new LimeExplainer(optimizedConfig);
    PredictionInput testPredictionInput = getTestInput();
    List<PredictionOutput> testPredictionOutputs = model.predictAsync(List.of(testPredictionInput)).get(Config.INSTANCE.getAsyncTimeout(), Config.INSTANCE.getAsyncTimeUnit());
    Prediction instance = new SimplePrediction(testPredictionInput, testPredictionOutputs.get(0));
    assertDoesNotThrow(() -> ValidationUtils.validateLocalSaliencyStability(model, instance, limeExplainer, 1, 0.5, 0.5));
}

Example 64 with LimeConfig

use of org.kie.kogito.explainability.local.lime.LimeConfig in project kogito-apps by kiegroup.

the class PrequalificationDmnLimeExplainerTest method testPrequalificationDMNExplanation.

@Test
void testPrequalificationDMNExplanation() throws ExecutionException, InterruptedException, TimeoutException {
    PredictionProvider model = getModel();
    PredictionInput predictionInput = getTestInput();
    Random random = new Random();
    PerturbationContext perturbationContext = new PerturbationContext(0L, random, 1);
    LimeConfig limeConfig = new LimeConfig().withSamples(10).withPerturbationContext(perturbationContext);
    LimeExplainer limeExplainer = new LimeExplainer(limeConfig);
    List<PredictionOutput> predictionOutputs = model.predictAsync(List.of(predictionInput)).get(Config.INSTANCE.getAsyncTimeout(), Config.INSTANCE.getAsyncTimeUnit());
    Prediction prediction = new SimplePrediction(predictionInput, predictionOutputs.get(0));
    Map<String, Saliency> saliencyMap = limeExplainer.explainAsync(prediction, model).get(Config.INSTANCE.getAsyncTimeout(), Config.INSTANCE.getAsyncTimeUnit());
    for (Saliency saliency : saliencyMap.values()) {
        assertNotNull(saliency);
        List<FeatureImportance> topFeatures = saliency.getTopFeatures(2);
        if (!topFeatures.isEmpty()) {
            assertThat(ExplainabilityMetrics.impactScore(model, prediction, topFeatures)).isPositive();
        }
    }
    assertDoesNotThrow(() -> ValidationUtils.validateLocalSaliencyStability(model, prediction, limeExplainer, 1, 0.3, 0.3));
    String decision = "LLPA";
    List<PredictionInput> inputs = new ArrayList<>();
    for (int n = 0; n < 10; n++) {
        inputs.add(new PredictionInput(DataUtils.perturbFeatures(predictionInput.getFeatures(), perturbationContext)));
    }
    DataDistribution distribution = new PredictionInputsDataDistribution(inputs);
    int k = 2;
    int chunkSize = 2;
    double f1 = ExplainabilityMetrics.getLocalSaliencyF1(decision, model, limeExplainer, distribution, k, chunkSize);
    AssertionsForClassTypes.assertThat(f1).isBetween(0.5d, 1d);
}

Example 65 with LimeConfig

use of org.kie.kogito.explainability.local.lime.LimeConfig in project kogito-apps by kiegroup.

the class PrequalificationDmnLimeExplainerTest method testExplanationStabilityWithOptimization.

@Test
void testExplanationStabilityWithOptimization() throws ExecutionException, InterruptedException, TimeoutException {
    PredictionProvider model = getModel();
    List<PredictionInput> samples = DmnTestUtils.randomPrequalificationInputs();
    List<PredictionOutput> predictionOutputs = model.predictAsync(samples.subList(0, 10)).get();
    List<Prediction> predictions = DataUtils.getPredictions(samples, predictionOutputs);
    long seed = 0;
    LimeConfigOptimizer limeConfigOptimizer = new LimeConfigOptimizer().withDeterministicExecution(true).withSampling(false).withStepCountLimit(30);
    Random random = new Random();
    LimeConfig initialConfig = new LimeConfig().withSamples(10).withPerturbationContext(new PerturbationContext(seed, random, 1));
    LimeConfig optimizedConfig = limeConfigOptimizer.optimize(initialConfig, predictions, model);
    assertThat(optimizedConfig).isNotSameAs(initialConfig);
    LimeExplainer limeExplainer = new LimeExplainer(optimizedConfig);
    PredictionInput testPredictionInput = getTestInput();
    List<PredictionOutput> testPredictionOutputs = model.predictAsync(List.of(testPredictionInput)).get(Config.INSTANCE.getAsyncTimeout(), Config.INSTANCE.getAsyncTimeUnit());
    Prediction instance = new SimplePrediction(testPredictionInput, testPredictionOutputs.get(0));
    assertDoesNotThrow(() -> ValidationUtils.validateLocalSaliencyStability(model, instance, limeExplainer, 1, 0.4, 0.4));
}