Instructor: Thierry Denoeux

Description: This course will present the main modern and classical methods of statistical computing, with examples from econometrics. The course will be composed of six chapters structured in two main parts introducing the foundations of statistical computing. Computer projects will be proposed using the R statistical software.

Program

1. Part I: Optimization
   1. Continuous optimization
   2. Combinatorial optimization
   3. EM algorithm
2. Part II: Simulation
   1. Classical simulation methods
   2. Markov-Chain Monte Carlo
   3. Bootstrap

Recommended reading: Computational Statistics book by Givens and Hoeting.

Slides

1. Continuous optimization
2. Combinatorial optimization
3. EM algorithm
4. Classical simulation of probability distributions
5. Markov Chain Monte Carlo methods
6. Bootstrapping

Resources

• R home page
• The art of R programming
• Applied Econometrics with R
• R Markdown cheat sheet

Exercises

1. Continuous optimization
   1. Solutions
2. Combinatorial optimization
   1. Solutions
3. EM algorithm
   1. Solutions
4. Classical simulation of probability distributions
   1. Solutions
5. Markov Chain Monte Carlo methods
   1. Solutions
6. Bootstrapping
   1. Solutions
7. Problem
   1. Solution

Data

1. Movie Buzz data
2. Transportation
3. F5_2 (macoeconomics dataset)
4. German health care data
5. Baseball
6. Mining disasters
7. Program effectiveness
8. Data_hmm.txt
9. Investment
10. Wages-productivity
11. Residential building (see description here)
12. Communities and crime