L2R

Researcher here (in ML), our field is so full of noisy results due to this issue. Everyone talks about it, but you can't get around the fact that you no longer can get away with a low amount of publications.

The issue is it takes away a significant amount of utility without adding any. Volume/brightness is slightly better, but you can get the same effect by holding down the physical button. Some users like to use Caps Lock for Ctrl, and this is an instance of the touchbar forcing a loss of utility with no gain.

Function keys are used heavily in IDEs, and forcing a large reach of fn + f# key isn't adding any utility either.

I have yet to see a situation where the touchbar adds some utility that merits the loss in other areas for users.

Based off of the results, you have to train a larger number of architectures to identify the right subnetwork.

There are approaches to ensure parameters remain stable despite the depth (selu, for example).

Nothing wrong with an SVM. How else would they create a decision boundary for classifying patients? The choice of the polynomial kernel is interesting, but I don't think it causes any issues given the data.

Spot on. Just a heads up, there is a decent amount of work on using convolutions to condense the initial representations and can reduce computation time equal to your max pooling. A lot of these tasks can be done via hyperparameter search over CNNs, so you can easily reach parity using a CNN-LSTM approach w/ the same number of parameters.