The Eighty Five Percent Rule for optimal learning
Wilson, Shenhav, Straccia, Cohen 2019 . (View on Nature →)
Researchers and educators have long wrestled with the question of how best to teach their clients be they humans, non-human animals or machines. Here, we examine the role of a single variable, the difficulty of training, on the rate of learning. In many situations we find that there is a sweet spot in which training is neither too easy nor too hard, and where learning progresses most quickly. We derive conditions for this sweet spot for a broad class of learning algorithms in the context of binary classification tasks. For all of these stochastic gradient-descent based learning algorithms, we find that the optimal error rate for training is around 15.87% or, conversely, that the optimal training accuracy is about 85%. We demonstrate the efficacy of this ‘Eighty Five Percent Rule’ for artificial neural networks used in AI and biologically plausible neural networks thought to describe animal learning.
Together with behaviour change, the ability to learn efficiently has always interested me. I’d heard of the Goldilocks learning zone before → that sweet spot of difficulty, not so hard that a problem is intractable and we lose motivation, not so easy that we’re going to succeed all the time and not progress. It’s interesting to see that appear in machine learning algorithms too. It aligns closely to what was seen in this study of classrooms.