Interpretable Narrative Explanation for ML Predictors with LP: A Case Study for XAI