• y- axis: probability
• colored dots: training data
• squiggle: fit of the data which tells us the probability that someone loves Troll 2

Define a threshold:

Probabilities vs Likelihood

It is good to know the difference between Likelihood and Probabilities (see StatQuest p. 112).

In short: Likelihood is the y-axis of a normal distribution. Probabilities are derived from a Normal Distribution by calculating the area under the curve between two points. The probability of getting any specific measurement is always 0 because the area of sth. with no width is 0.

We can fit a curve so that it maximizes the Likelihood (as show in this example):

Fitting a squiggle to data

As shown above we fit a squiggle to maximize Likelihood:

Above we see difference between logistic regression and linear regression.

Important: Calculating the Likelihood is different for people who Do Not Love Troll 2 because the y-axis is the probability that the Love Troll 2.