Abstract: Many decision professionals are not sure what to make of Machine Learning (ML). It intrigues us, yet we doubt its applicability to strategic decision making. The source of our intrigue could be that we enjoy working with models, and ML is an increasingly popular technique for creating models. Perhaps the source of our skepticism is that strategic decisions tend to be one-off events that involve a great deal of human engagement. This seems incompatible with ML models since they require large datasets and tend to automate actions.
The field of ML can be vast and overwhelming. This talk aims to demystify ML and discuss its usefulness for strategic decision making by exploring three statements about ML models: a) they are not all predictive b) they are imperfect c) they are not fit to be strategic decision models. We will discuss what makes each true, and what each implies for a decision professional.
Click on the file below to hear a sample of the presentation.