4 ms·
How well does "linear least squares to some nonlinear methods, simulated annealing, bayesian methods" work when you're doing speech to text, text generation or
by ipsum2 6y ago
How well does "linear least squares to some nonlinear methods, simulated annealing, bayesian methods" work when you're doing speech to text, text generation or object detection? Deep learning is different in that it's a huge leap closer to human capabilities compared to the other methods you've listed.
- f6v 6y agoDeep learning is closer to linear regression than it’s to human capabilities.
- mr_toad 6y ago> Deep learning is closer to linear regression than it’s to human capabilities. Quantitatively perhaps, in that deep learning is equivalent to many neurons, whereas a least squares model is only equivalent to a single neuron. Other than that there is not much evidence that a human brain is qualitatively different.
- f6v 6y agoDeep learning is much more than a "long" network nowadays. Most of DL success comes from clever architectures, like Inception Net. AFAIK there's no evidence human brain is anything like those fancy deep NNs.
- fock 6y agowell, your neural network is just a very complicated - ideally continously differentiable - function. In principle you could just fit this function to data with nonlinear least squares (e.g. Levenberg-Marquardt), we use backpropagation and SGD, because it's faster. Concerning simulated annealing: have a look at the old Alphafold and check how they optimize their neural networks... So, could you explain again how the concepts and foundations of deep learning differs from plain old regression techniques?