### Bayesian school

### Subjectivism, epistemic, logical probability

Bayesian modeling means that you can

**condition**an event. It sees the world as succession of random event.Frequentist Unlike the approach that only calculates the probability of data given a hypothesis, Bayesian modeling considers the posterior probability that a hypothesis is true given the data from the perspective of belief.

Frequentist believes there are some parameter behind randomness, while Bayesian solely rely on data considering everything is random with uncertainty. Prior is assumption based on dataset which determines posterior probability.

It is interpreted as the reliability of probability and is based on that.

#### Relative probability

The likelihood of an event occurring compared to another event, often expressed as a ratio or fraction.

#### PAC-Bayesian

It presents probabilistic bounds that guarantee the model's performance and focuses on calculating the generalization error of how well a given learning algorithm will perform using PAC