Statistics for ML
Deep dive into the whole required stastistics , that would be a required to learn stable diffusion from the very scratch Probability and Distributions Probability : It tells the chances of an event to happen. And is calculated as the ratio of no. of outcomes of a particular event to the total no. of outcomes. Likelihood : It measures how well a statistical model fits the observed data. Expectation : so E(x) is states what is the average value of x , now let say we have E(tau ~ P(Theta)) = {R(tau)}, this means what is the average reward value,given my tau comes from the distribution P(theta) so its expanded as SUM P(tau | theta) * R(tau) = Expectation {R(tau}, this is how this all plays out ! ...