Derive the mathematical relationship between cosine similarity and Eu- clidean distance when each data object has an L2 length of 1.

Here, we further explore the cosine and correlation measures.

Let x and y be two vectors where each vector has an L2 length of 1.

For such vectors, the variance is just n times the sum of its squared

attribute values and the correlation between the two vectors is their

dot product divided by n.