Statistics and machine learning: variations on a theme

David Jamieson Bolder

Economists and financial professionals face a broad range of old and new challenges. Economic systems are complex and constantly evolving in ways that are extremely difficult to predict. Corporate and economic policies must respond to this reality. The perhaps oldest, and most fundamental, question faced by economists and financial professionals thus relates to what is generally referred to as decision-making under uncertainty. More precisely, this involves determining a choice of action in the face of multiple alternatives and unknowns without any way of knowing in advance, or with certainty, the outcome of our choice. Such decisions come in many forms; a non-exhaustive list includes:

• • finding the appropriate company- and industry-specific tradeoff between risk and return;

• • determining a firm’s appropriate financing mix and capital structure;

• • deciding how to effectively spend, and target, a firm’s marketing and advertising budget;

• • identifying and structuring projects and investments that add to the value of the firm;

• • deciding on the course of monetary policy;

• • making alterations to a national