04. Probability Distributions

概率分布入门 1

What is a Probability Distribution?

Probability distributions allow you to represent the probability of an event using a mathematical equation. Like any mathematical equation:

  • probability distributions can be visualized using a graph especially in 2-dimensional cases.
  • probability distributions can be worked with using algebra, linear algebra and calculus.

These distributions make it much easier to understand and summarize the probability of a system whether that system be a coin flip experiment or the location of an autonomous vehicle.

概率分布入门 2

Types of Probability Distributions

Probability distributions are really helpful for understanding the probability of a system.
Looking at the big pictures, there are two types of probability distributions:

  • discrete probability distributions
  • continuous probability distributions

Before we get into the details about what discrete and continuous mean, take a look at these two visualizations below. The first image shows a discrete probability distribution and the second a continuous probability distribution. What is similar and what is different about each visualization?

Discrete Distribution (left) and Continuous Distribution (right).

Discrete Distribution (left) and Continuous Distribution (right).

Based on the visualizations, which of the following are true about the discrete probability distribution versus the continuous probability distribution?

SOLUTION:
  • The x-axis represents the main variable/event of interest for both visualizations.
  • In the discrete visualization, the x-axis variable can only take on certain values such as 1, 2 or 3.
  • In the continuous visualization, the x-axis variable can take on any real number value from -infinity to +infinity.

More terminology

  • Prior - a prior probability distribution of an uncertain quantity, such as the location of a self-driving car on a road. This is the probability distribution that would express one's beliefs about the car's location **before ** some sensor measurements or other evidence is taken into account.
  • Posterior - the probability distribution of an uncertain quantity, after some evidence (like sensor measurements) have been taken into account.

And you'll learn more about this terminology in the upcoming videos!