The question is pretty basic. My Erdos Number is number of edges that seperates me from Mr. Erdos. Two people are connected with an edge if they contributed a paper together. I had written a master thesis under guidance of my mentor. He is mentioned on the front page but technically he is not the second author. Can I use this as a link in calculating my Erdos Number?
Let me expand on user37208's comment. Nobody takes Erdős numbers too seriously, and the rules aren't quite precise. For example, do preprints count as lowering your Erdős number? Purists might say no, you have to wait until the paper is officially published, but nobody would judge you too harshly if you jumped the gun a little and counted a preprint.
However, everyone agrees on one basic rule: you can only count coauthored mathematics papers. Your thesis is not exactly a paper, and in any case your advisor is not a coauthor of your thesis, so it doesn't count. If you prepare a coauthored paper based on the thesis, then the situation will be different.
If you hang around with people who enjoy talking about Erdős numbers, you'll hear all sorts of weird variants. For example, people sometimes say that Hank Aaron has an Erdős number of 1 because he and Erdős once signed the same baseball, but this isn't intended seriously. Basically, if you can come up with a funny or surprising angle on things, people will let you bend the rules in all sorts of ways, but they won't really consider it legitimate.