Footnote:
Diese Datenquelle enthält auch Bestandsnachweise, die nicht zu einem Volltext führen.
Description:
Driven by the potential of digitalization, statistical learning and machine learning methods are commonly used for scheduling complex processes or forecasting in supply chain domains. However, trust in such methods is hampered by uncertainties in data quality, data exchange platforms, and data processing, affecting its traceability and reliability. Decision-relevant output provided by such methods is prone to trust issues in the data used for training, in the resulting model, and in the infrastructure in which the model is embedded. Considering the vulnerability of supply chains, wrong decisions have far-reaching consequences, raising the question of to what extent systems alone should be trusted for strategic, operational, and tactical decision-making. In this paper, we take a multidisciplinary perspective with the intention to analyze trust in statistical learning and machine learning methods from an organizational, sociological, and procedural perspective. The information base for this article is gathered through a systematic literature review. The central results of our research are a concept matrix comparing papers based on relevant criteria derived from literature and subsequent findings derived from this matrix. We encourage researchers in the fields of supply chain management, sociology, and statistics or machine learning to open up for interdisciplinary research and to build upon our findings.