academiamarginalia

The life and self-reflections of a professional mathematician

Category: statistics

Data Visualization

If I were in UK, I would go to London Transport Museum to see a new poster display showing historical versions of data visualization, Painting by Numbers. Very nice stuff.

By the way, speaking of transport, who in the world can tell me when the USC metro station is going to open?

Advertisements

Statistics in Cartoons

This is the first book on statistics I read from cover to cover! :-)

I have serious doubts it can serve as a textbook, and I would be really surprised if someone without any knowledge of statistics could really learn it from this book. Gonick&Smith‘s book is a great source of entertainment for those who love cartoons – the cartoonist is the first author for a good reason – and love statistics, and, therefore, know it already at some level. I would definitely recommend it to both beginners (as a complementary text for keeping motivation) and to professionals (as, for instance, a perfect in-flight reading).

There is just one tiny thing I am not comfortable with: throughout the book authors maintain the idea that it is ok to be afraid of mathematics. I completely understand why they do this, yet I feel this is a wrong message. To me, to be afraid of mathematics is the same as to be afraid of dogs. Yes, indeed, a dog bite in childhood may develop a life-long phobia of dogs. Similarly, a bad first math teacher may cause a “math allergy” for the rest of life. However, it is not good to be scared of either. Both mathematics and dogs are our friends. Big friends, indeed!

Loud Statisticians

The quiet statisticians have changed our world, not by discovering new facts or technical developments but by changing the ways we reason, experiment, and form our opinions — Ian Hackling.

Well, statisticians are not quiet any more!
Watch The Joy of Stats with Professor Hans Rosling.

MCMC Revolution in Reliability Engineering

As I have already mentioned,  I was invited to give a talk at the Southern California Probability Symposium 2001. After some thinking, I have decided to talk about the MCMC revolution that has happened over the last decade in the field of reliability engineering that lies at the boundary of engineering sciences and applied probability. This talk will be in the spirit of this paper by Persi Diaconis of Stanford.

Southern California Probability Symposiums have a reach history.
This is the schedule for SCPS-2011:

  • 9:50 – 10:40  Jinqiao Duan, IPAM and Illinois Institute of Technology, “Random Dynamical Systems with Non-Gaussian Noises”
  • 11:10 – 12:00  Tom Alberts, Cal Tech, “The Continuum Directed Random Polymer and the KPZ Universality Class”
  • 2:00 – 2:50  Allan Sly, UC Berkeley, “Asymptotic Learning on Social Networks”
  • 3:20 – 4:10  Tomoyuki Ichiba, UC Santa Barbara, “On collision of Brownian particles and applications”
  • 4:10 – 5:00  Konstantin Zuev, USC, “Markov Chain Monte Carlo Revolution in Reliability Engineering”
  • 6:00  Dinner

It is a great pleasure and honor for me to be in this company!

Accepted!

This paper has been accepted for publication in Computers and Structures.

This is the final comment from one of two reviewers:

Thank you very much for your answers on my comments and their consideration. I have no further remarks and appreciate this very interesting paper very much.

:-)

arXiv

The arXiv is a free scientific repository hosted by Cornell University, where physicists, mathematicians, and statisticians post their preprints. It was estimated that (as of September 2011) the repository had accumulated approximately 700,000 papers, with more than 6,000 new submissions arriving each month.

I have been using arXiv for a long time already, but, by some mysterious reason (a mixture of laziness and modesty), I did not put my own papers on arXiv. A couple of days ago I did my first bit by uploading my last paper co-authored with J.L. Beck of Caltech.  In this paper, we introduce a new scheme for Bayesian inference which is based on Markov chain Monte Carloimportance sampling, and simulated annealing.