What does the 'expected value' E represent in mean integrated squared error (MISE)

I am trying to use mean integrated squared error (MISE) to determine the optimal parameters for constructing a histogram. For now I construct histograms from samples of data and then compare them to their underlying probability density functions.

This is all going well, but I don’t understand what the E or ‘expected value’ given in MISE equations stands for. The value can be seen on the MISE wikipedia entry or more formally in kernel density estimation sources such as page 194 here.

In most published papers the expected value E is not defined at all, and the Wikipedia page is not particularly helpful for me. It seems to be the average value of something, but what? The histogram?

Thanks for any help,

M.