Treatment Regime (iii) 11/07/2007 Discussion for today - - PowerPoint PPT Presentation

Review of Tiastis’ slides on Dynamic Treatment Regime (iii)

11/07/2007

Discussion for today

* * 1 1 1

( , , ) { ( ) ( , 0) | , }

j j j j j j j j

l a E Y a Y a L l A a  

  

   

SA Murphy (JRSSB, 2003)

1 1 * 1 1 1 1 1 1 * 1

Dynamic programming: ( , ) sup { [ | , , ]}, ( , ) arg sup { [ | , , ]}, The subsequently for each , ( , ) sup { [ ( , ) | , , ]}, ( , ) arg sup { [

K K j j

K K a K K K K K K a K K K K j j j a K j j j j j j j j j a K

J L a E Y L a a d L a E Y L a a j J L a E J L a L a a d L a E J

           

   

1 1 1 * * * 1 2

( , ) | , , ]},......... Afterwards, the optimal rules are { , ,..., }.

j j j j j j K

L a L a a d d d

  

Too Naïve for a complex process