Section 1 of 24 · 4 lessons

Start with a Prediction

Suppose we know a person's height and want to estimate their weight. How would we find a useful rule, and how would we know whether it works for someone we haven't measured? We will start with that problem, then look at predictions that choose a category instead of a number.

0 of 4 finished
  1. 01Simple Linear RegressionTaller people tend to weigh more, but height does not determine weight exactly. We will use that relationship to build a prediction rule and work out what makes one line fit the observations better than another.
    Build with Botie →Read the lesson
  2. 02Held-Out EvaluationA rule can fit the examples it has already seen without being useful on new ones. We need to put some examples aside and use them to check the predictions after fitting.
    Build with Botie →Read the lesson
  3. 03Logistic RegressionSuppose our answer needs to be yes or no. We can first estimate a probability from the measurements, then decide how large that probability needs to be before we choose yes.
    Build with Botie →Read the lesson
  4. 04Judging a ClassifierA model can make different kinds of mistake, and they may have very different consequences. Counting correct answers alone can hide the errors that matter most for the task.
    Build with Botie →Read the lesson