Sufficient conditions for mutual independence

72 Views Asked by At

Take $n$ random variables $X_1,\ldots, X_n$ and suppose that $$ X_i\perp X_1,\ldots, X_{i-1}, X_{i+1},\ldots, X_n \quad \forall i \in \{1,\ldots,n\} \tag{$\star$} $$ Does this imply that $X_1,\ldots, X_n$ are mutually independent?

I believe yes but I want to double check with you. Here's my proof where $f_{\cdot}$ denotes probability distribution or pdf.

  • By $(\star)$ $X_1\perp (X_2,\ldots, X_n)$. Hence, $f_{X_1,\ldots, X_n} =f_{X_1}\times f_{X_2,\ldots, X_n}$

  • By $(\star)$ $X_2\perp (X_1,X_3,\ldots, X_n)$ which implies $X_2\perp (X_3,\ldots, X_n)$ Hence, $f_{X_1,\ldots, X_n}=f_{X_1}\times f_{X_2}\times f_{X_3,\ldots, X_n}$

  • By $(\star)$ $X_3\perp (X_1,X_2,X_4,\ldots, X_n)$ which implies $X_3\perp (X_4,\ldots, X_n)$ Hence, $f_{X_1,\ldots, X_n}=f_{X_1}\times f_{X_2}\times f_{X_3}\times f_{X_4,\ldots, X_n}$

  • ... $f_{X_1,\ldots, X_n}=\prod_{i=1}^n f_{X_i}$ which means mutual independence.