The Art and Craft of Systems Analysis (Part 1)

Both art and craft are involved in all major steps of systems analysis – from posing a specific research question, to specifying the assumptions underlying a model, to transforming those assumptions into mathematical or computational form, and to representing and communicating the results. These features are making systems analysis an exciting and creative process.

This session considers the creative tensions inherent in this process by touching on the following questions: What modern approaches can be recommended for model selection, calibration, and validation? How are systems boundaries best to be defined, emergent phenomena tackled, robustness checks designed, and validity limits delineated? How can multiple models be compared, interfaced, and integrated? How can a proliferation of false positives be avoided as the pool of data that can be correlated explodes? How can big data be used while skirting a temptation to construct big models in which a plethora of nonlinear interactions defies understanding and associated uncertainties cannot be assessed? In any of these regards, are best-practice recommendations becoming available that help delineate a safe operating space for systems analysis?


Introduction & chair: Prabhat Ranjan, TIFAC & IIASA Council Member for India

  • Mark Beaumont, University of Bristol: Approximate Bayesian computation: Methods and applications for complex systems – PDF
  • Werner Römisch, Humboldt Universität Berlin: Energy systems under uncertainty: Modeling and computations – PDF
  • Petr Shebalin, Russian Academy of Sciences: Linking probabilistic and deterministic forecast methods: systems analysis for predicting earthquakes – PDF


Bookmark and Share

Location: IIASA Date: 11th November 2015 Time: 16:15 - 17:45 pr_26_8_14 Prabhat Ranjan mark beaumont Mark Beaumont Werner Roemisch Werner Römisch shebalin2 Petr Shebalin