#### worse than the Drake eq.

I was reading the June 2009 IMS bulletin on my way to Korea for the 1st IMS-APRM meeting. Then, I was in half shock and in half sadness. Something unlike than the Drake equation had happened.

The Drake eq. is used as an indicator that the chance of finding an organic society equivalent to the human society. As you guess, such chance is extremely low. What would be a chance that two obituaries of eminent statisticians who influenced many can appear in the same bulletin. Personal thought led that the obituary section of the bulletin is further extreme than the Drake equation.

If you are an astronomer who are interested in spectral analysis and looked for statistical or data analysis literature, you cannot miss I.J.Good’s **bump hunting** paper.

Density Estimation and Bump-Hunting by the Penalized Likelihood Method Exemplified by Scattering and Meteorite Data

by I.J.Good and R.A. Gaskins in JASA, Vol.75, No. 369, pp. 42-56

The penalized likelihood approach for density estimation and bump hunting and its Bayesian interpretation has popularized statistical application to spectrum type natural science data.

Not by the popularity but by my personal interest in computational geometry and its statistical expansion, Worlsey’s publications became my reading list. Computational geometry pertains the goodness of nonparametric statistics for multivariate data which are not well explored compared to nonparametric methods for univariate data. His introductory paper about computational geometry like

Keith Worsley (1996)

The Geometry of Random images (zipped postscript), Chance, 9(1), pp.27-40

can be informative and useful to some astronomers.

Speaking of the Drake equation, it was the first thing that gave me a notion of probability, it describes how one would simply formulate and compute the chance of finding life beyond the earth. The equation is a process of constructing a likelihood function. In fact, I didn’t think this equation to be a likelihood function at that time but its unique creativity carved my memory. The way this equation describes how to compute the chance of the existence of extraterrestrial intelligence is a good example of **chain rule** in modifying likelihood functions.

I have never met those scholars face to face but through their writings, their works shaped my way of thinking. This personal experience made me hard to believe obituaries of two respectful statisticians. It was like getting estimates of the chance of meeting ETs which I found very small when I played with the equation. Although their chances are extremely low, things can happen. Finding life outside of the earth and finding a sad news of two eminent scientists’ death are alike.

## Leave a comment