Your Market Can't Learn XOR
January 8, 2018 7:00 AM Subscribe
Your Market Can't Learn XOR
Math for the Arrow-Debreu model of markets and the perceptron model of neurons is remarkably similar, down to similar derivations for linear equilibrium via linear separability in solution space. I talk about consequences.
Math for the Arrow-Debreu model of markets and the perceptron model of neurons is remarkably similar, down to similar derivations for linear equilibrium via linear separability in solution space. I talk about consequences.
Role: did everything
« Older London charity shops treasure hunt.... | arreXted developMEN... Newer »
...though I do have a suspicion that "when all you have is a hammer, everything looks like a nail" explains the equations arrived at for both neural networks and economic models.
Still, it's fun to think about what hidden layers might map to in an economy, or how stochastic gradient descent might work.
posted by clawsoon at 6:11 AM on January 12, 2018