2004-2005

Please check out this page for regular updates and classnotes.

• Teacher: Arnab Chakraborty
  • Email: arnab_v@isical.ac.in
  • Office: Room 5.8, fifth floor, new building.
• Class room: 527
• Class time:
  • Monday, 10:15--12:00
  • Wednesday, 10:15--12:00
  • Thursday, 10:15--11:05
• Weight distribution:
  • Assignment: 20
  • Midterm: 30
  • Semestral: 50

1. Classical nonparametric methods:
   • Standard paired sample, one sample, two sample and k-sample methods.
   • U-statistics
   • Properties of tests
2. Computer-intensive methods:
   • Bootstrap
   • Cross-validation and CART
   • Jackknife
3. Sequential inference:
   • Dynamic programming
   • Sequential testing and SPRT
   • Wald's equations, Wald's approximation and fundamental identity

Reference and textbooks etc

• All the classnotes are online. These will constitute your main reference material. The lecture notes contain many exercises which will be your assignment problems.
• For the classical nonparametric methods the book by Hollander and Wolfe is good. It is a fat book, but we shall need only some parts of it.
• The book by Gibbons and Chakraborty is also a standard reference.
• For computer-intensive nonparametric methods the book by Efron and Tibshirani is a nice book to read.
• There are many excellent books on sequential statistics, however almost all of them use measure theoretic notations. So they may not be easily accessible to you. Here are a few standard books on the subject:
  1. Sequntial statistics by Wald: This is a very old book, written by one of the fathers of the subject. It is written informally, makes easy reading, but does not contain modern treatments.
  2. Sequential statistics by Siegmund: Good modern book. But may not be easily accessible to you at this stage.
  3. Great Expectations by Chow , Robbins and Siegmund: Another great book, but makes heavy use of measure theory.

Anonymous suggestions