Maximum Likelyhood distro
Esentially trying to derive the function from several argument inputs. (What maxamizes the likelyhood function). Using data to come up with a model & its params (whr original model is unknown).
Intro:
Coin flip? Four times in a row pops up heads? Likelyhood this could happen?
Wrong line of thought: Would think but ! would think 50% chance heads on each, then sum them tgt → . This occurs by thinking each event is independent (one event does ! affect other), and identically distributed (means each prob is a half).
Likelyhood func: likelyhood of obtaining set of observations as func of @param of a model
Originally we assumed that it was 50/50 for heads/tails. ! know if fairly weighted. So how would we find out the actual prob (if actually 50/50 chance). We assign heads to be p and tails 1-p (the rest). So we culd plot frm where if means either 100% heads or 100% tails. So, to answer the original qn, with our input of four times being heads, the answer that would maxmize this would be x=1, or that heads is always one (heads).