Where Do Loss Functions Come From?

We all know that in Linear Regression we aim to minimise the Sum of Squares Error (SSE) as our objective. However, why is it the SSE and where does this expression even come from? In this article I hope to answer this question using something called the Maximum Likelihood Estimator. Spending enough time in the Data Science community I am confident you would have come across the term Maximum Likelihood Estimator (MLE). I am not going to give a super in detail analysis of MLE, primarily because it has been done so many times in different ways that are probably better than I could ever explain it.