Machine Learning With Go

Hypothesis testing

We can quantify how likely questions with probabilities and can even calculate conditional probabilities with Bayes theorem, but how can we quantify how significant questions that correspond to real-world observations? For example, we can quantify the probability of heads/tails with a fair coin, but how can we determine how significant it is when we flip a coin a bunch of times and observe 48% heads and 52% tails? Is this significant? Does it mean that we have an unfair coin?

These how significant questions can be answered using a process called hypothesis testing. This process generally includes the following steps:

1. Formulate a null hypothesis, referred to as H₀, and an alternate hypothesis, referred to as H_a. H₀ represents ...