Welcome
I currently work as a data scientist at the Swiss Federal Railways. The content of this website stems mostly from my earlier life an academic at UC San Diego (CA) and ETH Zurich. That work revolved around statistial analysis of large volumes of geophysical data from widely distributed sensor networks. My last focus area was on combining graph theoretical approaches to studying sensor arrays in unknown media (read more on that in Network clustering in a distributed sensor array).
Some other scientific projects I’ve worked on in the past:
- Tracking human mobility across a geophone network
- Network clustering in a distributed sensor array
- Time-lapse study of rock anisotropy beneath underground gas storage
- Hydrocarbon microtremor statistics: the confounding variable issue
Methodologically, I am interested in (sparse) dictionary learning, (non-linear) high-dimensional dimensionality analysis, and more broady on what the possibilities and limits of statistical thinking are in the world we live in. As for the use of such techniques I’d like to conclude with two quotes on the nature of method more generally:
“What we observe is not nature itself, but nature exposed to our method of questioning.”
Werner Heisenberg, in “Physics and Philosophy” (Harper, 1958)
“The trick to being a scientist is to be open to using a wide variety of tools.” Leo Breiman, in “Statistical Modeling: The Two Cultures” (2001)
I’ve left acedemic research in 2016, but have been involved in various applied R&D-type projects in the industry since then.
About the author
I was trained as a physicist at ETH Zurich (Switzerland) with an emphasis on radio astronomy and acoustics. After a stint at the family-run travel agency (writing database applications and travel itineraries) I’ve joined the geophysical service company Spectraseis in research and developemt from 2006 to 2010. This led to an industry co-sponsored PhD project in 2010 at the ETH Zurich (with the Low Frequency Seismic Partnership consortium) to study the statistics of the ambient seismic background wave field. In 2014 I was awarded a Swiss National Science Foundation (SNF) fellowship to pursue postdoctoral research at UC San Diego (California), collaborating with Peter Gerstoft. I’ve worked on methods to analyze large sensor networks with minimal knowledge of medium properties using, among other things, Machine Learning techniques and graph theory approaches.
Mini CV
- since Jun 2022 – Lead Data scientist Sanitas AG
- Data applications and MLOps, analytics, AI strategy.
- Dec 2016-May 2022 – Senior Data scientist SBB
- Developing data applications, exploratory data anlysis and coordinating the SBB Data Science Community.
- May-Nov 2016 – Data scientist IMSD Sarl, Zurich
- Developing analysis and reporting solutions for the private sector.
- 2014-2016 – Postdoc UC San Diego
- Using machine learning tools to study wave fields measured by dense sensor networks.
- 2010-2013 – PhD geophysics, ETH Zurich
- Industry-academia co-sponsored project on ambient seismic statistics. Focus on spatial and polarization statistics and three-component array processing and its use for anisotropy measurement.
- 2007-2010 – Research scientist and software developer with Spectraseis
- Prototyping of processing and analysis algorithms, design and maintenance of production database and processing system, technical presentations to clients, software development. Spectraseis is a small geophysical service company.
- 2005-2006 – Database developer and allrounder at Riahi Travel,
- Riahi Travel is the family-run travel agency where I’ve spent much of my student time :-)
- 2000-2004 – MSc physics, ETH Zurich
- Including a student project as part of the Compact Muon Solenoid (CMS) experiment at the European Center for Particle Physics CERN and a measurement session at the ETH Zurich training radio telescope Apraxos.
nimariahizrh at gmail dot com
Site created using rmarkdown in RStudio. Last update: Jan 2020