Don’t let the perfect be the enemy of the good! Renouncing an approach that would be able to provide perfect results under ideal conditions for a simpler alternative can improve performance under real-world circumstances | |
There is no golden bullet! The optimal strategy cannot be identified in general but depends on preconditions of the specific use case | |
The diversity of epistemic approaches is an asset The above insights can be generalized beyond tailoring machine learning algorithms to a broader range of forecasting and decision making strategies. Examples for such alternatives range from expert intuition (“gut feeling” [11]) and deliberately simple decision rules (fast-and-frugal heuristics [9]), to debiasing subjective judgements (“crowd intelligence” [10, 12]) and the strategic integration across several sources of information and knowledge |