Tip:
Highlight text to annotate it
X
Let's now talk about a 2nd class of unsupervised learning avenues
that are called dimensionality reduction.
We're going to start with a little quiz, in which I will check your intuition.
Suppose we're given a 2-dimensional data field, and our data lines up as follows.
My quiz is: How many dimensions do we really need?
The key is the word really,
which means we're willing to tolerate a certain amount of error in accuracy,
because we're going to capture the essence of the problem.