In this unit you will learn about a parameter estimation strategy called maximum likelihood. This strategy writes down a function in terms of data and weights for that data, defines a target function—the likelihood score—and then works to iteratively choose values for the weights that literally maximize the likelihood of observing the data.