STAT 4444 Study Guide - Final Guide: Posterior Probability, Normalizing Constant, Prior Probability

• School: Virginia Tech
• Department: Statistics
• Course: STAT 4444
• Professor: Hamdy Mahmoud

Document Summary

Class business: homework iii is due on april 5 at 11:59 p. m. Let"s recall the example from the beginning of the semester when we considered what proportion of students will quit school given a 10% increase in tuition. Assume we assign the following prior distribution. p( ) = Now, let us nd the posterior distribution of . Remember that out of 50 students surveyed, 7 said that they would quit. p( |y) p( )l( |y) p( |y) . 1 e 1 e 7(1 )43 p( |y) e 7(1 )43. 0 e 7(1 )43d = c then the posterior distribution is as follows p( |y) = After nding the posterior distribution, you can integrate to nd the mean. To nd the predictive probability distribution, again use integration. p(y |y) = z 1. 0 p( |y) p(y | )d , y = 0, 1, , n. The simplest numeric algorithm for one-dimensional integrals is the composite midpoint rule.