ksvanhorn.com

Home Bayes Home Jaynes Errata Articles Books Software Contact

Chapter 13: Decision theory, historical background p. 403, line after equation (13.9): Insert a minus sign before the summation symbol in the definition of $I(q;p)$. p. 403, footnote 1: Arguably, ``predicted nonstorm'' should be ``nonpredicted nonstorm,'' as no storm occurred, contrary to predictions. p. 411, equation (13.18): A couple of small technical corrections. `` $\partial L / \partial \beta = 0$'' should be `` $f(x \mid \alpha) \partial L / \partial \beta = 0$''. The sentence that follows the equation is incorrect -- the equation really says that $\beta(x)$ must be a stationary point of $L$, regardless of $x$, which leads to the useless solution mentioned later on ( $\beta(x) = \alpha$).