Jan Galkowski

I haven’t been very active at *Azimuth* of late, both because of a great deal of study of new methods and techniques outside of and for work at *Akamai* *Technologies* of late, and because of a recovery period from the Azimuth Data Backup Project which was a bit of a *burnout* for me, even though I am delighted it was done and really enjoyed working with the team that made it happen. I’m also deeply grateful to Professor John Baez for throwing his support behind it and pushing it on. All indications are that our fears were justified but also that the Trump administration is either too ineffective or too corrupt (`A New Argentina`

) to be potent at environmental destruction.

I am a statistician and quantitative engineer working for *Akamai Technologies* in Cambridge, MA, where much of my time is devoted to customer-sponsored research regarding the Internet, and where I will continue to work for 5 or 6 more years before “retiring.” I live in Westwood, MA, with my wife, Claire, in a nearly zero Carbon-using home.

I’m an active student of environmental sciences, Bayesian methods, and of the data sciences. You can learn more about me here, and from my LinkedIn profile. I am a member of the American Association for the Advancement of Science, the Ecological Society of America, the American Statistical Association (active in its Boston Chapter), the International Society for Bayesian Analysis, and three organizations at Woods Hole Oceanographic Institution, the Associates, the 1930 Society, and its Fye Society.

I am also active in social and political activities relating to the environment, partly through the Green Congregation Committee at the Unitarian Universalist congregation to which I belong, First Parish in Needham, MA, and partly as a staunch advocate for distributed, locally owned solar energy. Some might call me a *solar* *revolutionary*, in the spirit of the late German parliamentarian, Hermann Scheer and John Farrell of the Institute for Local Self Reliance. In any case, I see solar energy as part of a democratization of energy and, so, political power: Take back control of your energy, and thereby take back control of your democracy.

“Warming slowdown? (part 1 of 2)” The idea of a *global warming slowdown* or *hiatus* is critically examined, emphasizing the literature, the datasets, and means and methods for telling such. Also available at the Azimuth Project wiki.

“Warming slowdown? (part 2 of 2)” The idea of a *global warming slowdown* or *hiatus* is critically examined, emphasizing the literature, the datasets, and means and methods for telling such.).

“Bayesian inversion of commingled tonnage of municipal solid waste to isolate components” Bayesian inversion to recover latent components in mixtures is a standard technique, with wide application. Yet, apparently, it is not well known. Frequentist methods for doing this are known as algorithms for *blind source separation*.

Unrelated to *Azimuth*, a technical and, occasionally, political blog which records developments in renewable energy, offers the occasional statistical and computational illustration and guidance, comments on Climate Science and other sciences of interest to me, e.g., Quantitative Population Biology, and sometime deep dives into subjects far from any of these, e.g., gun control or Geology.

- Statistical applications in quantitative Ecology and Population Biology, particularly examining dynamics of invasions for sessile species.
- Bayesian statistics, especially computational challenges, tutorials, and learning how to harness massive concurrency in its service.
- Applications of
*Generalized**Linear**Mixed**Models*using Markov Chain Monte Carlo methods. (See also this*Azimuth*introduction. - Boosting as a statistical technique, and its applications, particularly as seen as one of a variety of fascinating new methods for non-parametric- or model-free forecasting, and equation-free mechanistic forecasting.
- Developing means to interpret and validate models deviced using
*recurrent**neural**networks* - Statistical support to citizen science efforts, after Kosmala, Wiggins, Swanson, and Simmons.
- Mastering compelling graphics and illustrations using the facilities in LaTeX PGF/Tikz

Clearly, the rigor of statistical methods is something to which all practitioners should aspire, but there are practical challenges posed by some problems, such as dataset size or character, which make it difficult to apply traditional methods. There is also the key insight that if the primary interest is *prediction*, in the most general sense of the term, then to insist upon transparency and interpretability of the resulting model or mechanism might foreclose on the possibility of achieving the best possible or at least a more powerful predictive method. By *prediction* I mean precalculating a *response* given values for instrumental variables, whether or not the response has yet been observed. Presumably, methods are calibrating using paired sets of *responses* and *predictors*. *Forecasting* is generally restricted to cases where one of the predictors is time or a proxy for time.

The insight really isn’t that deep, for it simply follows from the observation that the set of all means and methods for prediction which are highly successful at prediction is *no* *smaller* *than* the set of all means and methods which are constrained to be interpretable or transparent to human inspection and still predict successfully. Given that the former set is in general larger than the latter, if something like a uniformly most powerful predictive mechanism exists in the former set, a set which necessarily includes the latter as a subset, it is more likely to be in the complement to the latter than in the latter. I believe *uniformly* *most* *powerful* could be rigorously formulated here in terms of a *loss* *function* and its *argmin* over all predictive mechanisms in the set of all.

- Inferring latent causes of shortages in the drinking water supply in the town of Sharon, MA and, more generally, including predictive inference.
- Developing techniques which facilitate interpretation of data gathered by volunteers in the field and natural settings censored by seasonal availability and varying quality.
- Studying ecosystem relationships of the much maligned
*Alliaria**petiolata*(“Garlic mustard”) considering the insights of Professor Peter Del Tredici and colleagues. See here for an incomplete overview.

jan@westwood-statistical-studios.org bayesianlogic.1@gmail.com

category: members