We now extremize S subject to two constraints, and so we define \[S^*\big(\{p\ns_n\},\lambda\ns_0,\lambda\ns_1\big)=-\sum_n p\ns_n \ln p\ns_n - \lambda\ns_0\Big(\sum_n p\ns_n -1 \Big) -\lambda\ns_...We now extremize S subject to two constraints, and so we define S∗({p\nsn},λ\ns0,λ\ns1)=−∑np\nsnlnp\nsn−λ\ns0(∑np\nsn−1)−λ\ns1(∑nX\nsnp\nsn−X). We then have \pzS∗\pzp\nsn=−(lnp\nsn+1+λ\ns0+λ\ns1X\nsn)=0, which yields the two-parameter distribution p\nsn=e−(1+λ\ns0)e−λ\ns1X\nsn. To fully determine the distribution \…
