Why does tSNE includes "Stochastic" on its name?

Question

I know that in Machine Learning one classification of algorithms that researchers use is if they are deterministic or stochastic. I've been studying tSNE, but I don't get if "Stochastic" is included in its name because there are hidden theoretic justifications that involves Stochastic Processes or because it uses probability theory by changing distances to probabilities and then using de Cauchy distributions to fix the crowding problem, etc.

Answer 1

I assume the term 'Stochastic' comes from the way the embedding is performed. This is a FAQ from the authors webpage:

Every time I run t-SNE, I get a (slightly) different result?

In contrast to, e.g., PCA, t-SNE has a non-convex objective function. The objective function is minimized using a gradient descent optimization that is initiated randomly. As a result, it is possible that different runs give you different solutions. Notice that it is perfectly fine to run t-SNE a number of times (with the same data and parameters), and to select the visualization with the lowest value of the objective function as your final visualization.

This is explained in more detail in section 2 of the original paper.