Explainable AI for Cost Estimation
For the application of complex machine learning models, explanationsand interpretability often play a crucial role besides accuracy. This isalso the case in Cost Estimation, where the employment is often viewedsceptically due to the lacking credibility of models (Smith and Mason,1997, pp. 21-23). Thereby, questions arise like: How to capture a model'sinterpretability? How does interpretability cohere with accuracy? In whichway contributes the research field of Explainable AI to those problems ingeneral? This paper aims to answer these questions with an analysis ofinterpretability in the literature, a simulated use case for cost estimationand an empirical survey. Finally, an approach for a twofold evaluationof both model accuracy and interpretability for cost estimation models isproposed.