No, I just thought about it some more, and I realized that increasing the learning rate of a model (assuming the optimizer is something like SGD) would inject more randomness, just like increasing the temperature of simulated annealing would.
No, I just thought about it some more, and I realized that increasing the learning rate of a model (assuming the optimizer is something like SGD) would inject more randomness, just like increasing the temperature of simulated annealing would.