Yasha Pushak

Contrastive Explanations For Machine Learning Forecasting Models

WARM START FOR MULTIPLIER TUNING POSTPROCESSING FOR MACHINE LEARNING BIAS MITIGATION

MITIGATING BIAS IN MACHINE LEARNING WITHOUT POSITIVE OUTCOME RATE REGRESSIONS

FAIRNESS FEATURE IMPORTANCE: UNDERSTANDING AND MITIGATING UNJUSTIFIABLE BIAS IN MACHINE LEARNING MODELS

MULTIPLIER TUNING POSTPROCESSING FOR MACHINE LEARNING BIAS MITIGATION

ACCELERATING AUTOMATED ALGORITHM CONFIGURATION USING HISTORICAL PERFORMANCE DATA

THRESHOLD TUNING FOR IMBALANCED MULTI-CLASS CLASSIFICATION MODELS

AUTOMLX COUNTERFACTUAL EXPLAINER (ACE)

LEARNING HYPER-PARAMETER SCALING MODELS FOR UNSUPERVISED ANOMALY DETECTION

UNIFY95: META-LEARNING CONTAMINATION THRESHOLDS FROM UNIFIED ANOMALY SCORES

Expert-optimal correlation: contamination factor identification for unsupervised anomaly detection

Fast and accurate anomaly detection explanations with forward-backward feature importance

N-1 EXPERTS: MODEL SELECTION FOR UNSUPERVISED ANOMALY DETECTION

AUTOMATED DATASET DRIFT DETECTION

LOCAL PERMUTATION IMPORTANCE: A STABLE, LINEAR-TIME LOCAL MACHINE LEARNING FEATURE ATTRIBUTOR

DATASET-FREE, APPROXIMATE MARGINAL PERTURBATION-BASED FEATURE ATTRIBUTIONS

EFFICIENT AND ACCURATE REGIONAL EXPLANATION TECHNIQUE FOR NLP MODELS

Fast, approximate conditional distribution sampling

Global, model-agnostic machine learning explanation technique for textual data

POST-HOC EXPLANATION OF MACHINE LEARNING MODELS USING GENERATIVE ADVERSARIAL NETWORKS

USING GENERATIVE ADVERSARIAL NETWORKS TO CONSTRUCT REALISTIC COUNTERFACTUAL EXPLANATIONS FOR MACHINE LEARNING MODELS

Generalized expectation maximization for semi-supervised learning