2
You have two questions in one.
Is it maxpool that ruins the model?
I would say no, the maxpool is a standard operation for convolution networks, it down-samples the intermediate representation to reduce the necessary computations, improve the regularization, and adds translation invariance to some degree. Originally averaging was used to downsample over ...
2
Why the number of training points to densely cover the space grows exponentially with the dimension?
First, let's try to build some intuition for what we mean when we say that we want to "densely cover" a $d$-dimensional space $\mathbb{R}^d$ of real numbers. For simplicity, let's assume that all values in all dimensions are restricted to lie in $[0, 1]$. Even with just a single dimension $d=1$, there are actually already infinitely many different ...
1
Accuracy is a good measure if our classes are evenly split, but is very misleading if we have imbalanced classes.Always use caution with accuracy. You need to know the distribution of the classes to know how to interpret the value.
Only top voted, non community-wiki answers of a minimum length are eligible
