“These many models are actually better than individual models, even if the many models are all equally good on their own. So, you get the wisdom of crowd’s effect being the fact that the diversity of models is really powerful. That’s why they call this the Diversity Prediction Theorem because diversity ends up being really valuable in crowds of people getting things right. Alright, so that’s an outline of what we’re gonna do. We’re gonna talk about prediction at individual level and at the collective level and we’ll show how having many models results in something what you call [inaudible] prediction theorem and then diverse sets of models need to more accurate collective forecast and then explains what some people call a wisdom of crowds phenomena.” – Transcript from Scott Page Coursera
Source:
Scott Page Model Thinking MOOC Course
