Importance Sampling
Last edited: August 8, 2025Key insight: suppose you have some fairly rare event and you want the likelihood of it. We can do this by drawing normal samples and reweighing them.
Suppose we want \(p_{\text{fail}}\); and we have \(q\) the proposal distribution and \(p\) the nominal distribution:
\(\tau \sim q\qty(\cdot)\), \(p_{\text{fail}} = \int 1 \qty {\tau \not\in \psi} p\qty(\tau) \dd{\tau }\)
What if we define a weird \(1\) such that:
\begin{equation} 1 = \frac{q\qty(\tau)}{q\qty(\tau)} \end{equation}
inclusion exclusion counting
Last edited: August 8, 2025If an outcome can be from sets \(A=m\) or \(B=n\) with no overlaps, where \(A \cap B = \emptyset\), then, the total number of outcomes are \(|A| + |B| = m+n\)
If there are overlap:
\begin{equation} N = |A|+|B| - |A \cap B| \end{equation}
independently and identically distributed
Last edited: August 8, 2025\(n\) random random variables are IID if they are
- independent
- identically distributed (see below)
“identically distributed”
Consider \(n\) random variables:
- \(X_i\) all have the same PMF / PDF
- and therefore, all have the same expectation and variance
central limit theorem
when things are IID, you can use central limit theorem.
Index Index
Last edited: August 8, 2025Here’s a list of all indexes:
This should be reflected on a fancier way on my home page.
