We can implement this algorithm to find the posterior distribution P(A|B) given the product of P(B|A) and P(A), without considering the normalizing constant that requires a complex integration.
Let's suppose that:
Therefore, the resulting g(x) is:
To solve this problem, we adopt the random walk Metropolis-Hastings, which consists of choosing q ∼ Normal(μ=x(t-1)). This choice allows simplifying the value α, because the two terms q(x(t-1)|x') and q(x'|x(t-1)) are equal (thanks to the symmetry around the ...
