1st Edition
Sensitivity Analysis for Business, Technology, and Policymaking Made Easy with Simulation Decomposition (SimDec)
SimDec is a revolution in decision-making support. SimDec “teases out” inherent cause-and-effect relationships and reveals the intricacy of relationships between sets of input and output variables. At its core, SimDec is an amalgamation of uncertainty and global sensitivity analysis with an innovative visualization technique. While straightforward and elegant, this novel approach significantly enhances the analytical capabilities of users by readily exposing seemingly, a priori, counter-intuitive behaviors so that they can be readily understood by both technical specialists and non-technical users, alike.
This book is the first to articulate the ubiquitous applicability of SimDec and has been written by the leading proponents of the technique. The book provides the necessary background to fully understand the underlying approach and then demonstrates its applicability to a wide spectrum of fields such as finance, entrepreneurship, energy, 3D manufacturing, geology, the environment, engineering, public policy, and even superconducting magnets. To facilitate as widespread adoption and penetration of SimDec as possible, all supporting computer codes are available, open-source, in Python, Julia, R, and Matlab.
The innovative material will be of primary benefit to practitioners and researchers analyzing data from the social sciences, business, science, engineering, mathematics, and computing.
1. Methodological landscape of sensitivity analysis and the place of SimDec
Mariia Kozlova, Samuele Lo Piano, and Julian Scott Yeomans
2. SimDec algorithm and guidelines for its usage and interpretation
Mariia Kozlova, Robert J. Moss, Pamphile Roy, Abid Alam, and Julian Scott Yeomans
3. Overview of SimDec applications
Mariia Kozlova and Julian Scott Yeomans
4. Unlocking actionability in financial modeling with Simulation Decomposition
Roman Stepanov, Mariia Kozlova, and Julian Scott Yeomans
5. Unpacking the role of contextual factors in public support for mitigating revenue risk in public–private partnership projects
Roberta Pellegrino, Mariia Kozlova, Luiz Brandao, and Julian Scott Yeomans
6. Printing homes: Unit cost estimation for additive manufacturing in construction
Alexander N. Walzer, Mariia Kozlova, and Julian Scott Yeomans
7. Where should we go? Deep tech market entry decisions through the lens of uncertainty
Alexander Myers, Mariia Kozlova, and Julian Scott Yeomans
8. Uncertainty considerations in life cycle assessment of COVID-19 masks: single-use versus reusable
Natalia Vinitskaia, Anna Zaikova, Mariia Kozlova, and Julian Scott Yeomans
9. Model-fidelity analysis for sequential decision-making systems using simulation decomposition: Case study of critical mineral exploration
Robert J. Moss, Mariia Kozlova, Anthony Corso, and Jef Caers
10. Upgrading the toolbox of techno-economic assessment with SimDec: Power-to-X case
Hannu Karjunen, Sini-Kaisu Kinnunen, Arto Laari, Antero Tervonen, Petteri Laaksonen, Mariia Kozlova, and Julian Scott Yeomans
11. Capturing multi-dimensional nonlinear behaviour of a steel structure reliability model – global sensitivity analysis
Antti Ahola, Mariia Kozlova, and Julian Scott Yeomans
12. Sensitivity analysis of a superconducting magnet design model
Manuel García Pérez, Mariia Kozlova, Susana Izquierdo Bermúdez, Juan Carlos Pérez, and Julian Scott Yeomans
13. New level of personal decision-making: day-to-day choices with SimDec
Anna Sidorenko, Daria Moshkivska, Mariia Kozlova, and Julian Scott Yeomans
Epilogue
Afterword. SimDec meets SIPmath
Sam L. Savage
Biography
Mariia Kozlova is a Post-Doctoral Researcher at LUT University Business School, Finland, and a Visiting Scholar at Stanford University, USA.
Julian Scott Yeomans is Professor and Director for the MMAI and MBAN programs at the Schulich School of Business, York University, Canada.
"This book on sensitivity analysis by Mariia Kozlova and Julian Scott Yeomans summarizes the state-of-the-art method for computational model analysis, the revolutionary Simulation Decomposition (SimDec). For readers like me working in the general areas of modelling, optimization and machine learning, I find this book extremely useful because it essentially has everything about SimDec in one package, which makes it an effective tool to do global sensitivity analysis. The results are easy to understand, with colorful visualization, uncertainty quantification, and a wide spectrum of diverse applications. In addition, the open-source SimDec software with code packages in Python, R, Julia and Matlab will revolutionize the ways for training next-generation scientists and practitioners to do the right kind of sensitivity analysis so as to figure out the most influential factors correctly and to support more informed decision-making."
Xin-She Yang, Reader at Middlesex University London, Fellow of the Institute of Mathematics and its Applications (FIMA), UK
"Simulation decomposition is an incredibly powerful technique that allows researchers and engineers to identify key factors influencing the performance of a system and make targeted improvements. This book provides a simple and accessible guide to understand and apply the technique to practical problems. The accompanied open-source SimDec code is the key to enable the reader to quickly learn and apply the method."
Leifur Leifsson, Associate Professor, School of Aeronautics and Astronautics, Purdue University, USA
"Simulation Decomposition methodology, a “Monte Carlo”- based computational algorithm, is quickly becoming a game-changer in the world of engineering, industry, and finance. In this newly published book, Julian Scott Yeomans and Mariia Kozlova explore the importance of this method in providing an accurate and detailed holistic picture of the behavior of complex systems. The book delves into the real-life applications of Simulation Decomposition, highlighting its effectiveness in optimizing processes and improving product designs. Through detailed case studies and insights from industry experts, readers will gain a thorough understanding of this powerful methodology and its potential for transforming the way we approach complex systems."
Kambiz Vatan-Abadi, Chief Innovation Officer, CI Financial, Canada
"SimDec is an ingenious way to tame the combinatorial complexity of systems’ behavior in real-world decision making. Its deceptively simple approach visually reveals the hidden factors that materially impact an uncertain outcome. This new method for determining sensitivity indices seems a hidden gem. Its transparency is invaluable to practitioners. Credit is also due to the authors for making an otherwise challenging topic most entertaining and accessible to practitioners. SimDec’s ease-of-use should make it the de facto standard for data analysis in industrial, engineering, and scientific environments."
Kalyan Moy Gupta PhD, Founder and Vice President of Research, Knexus Research, Washington DC, USA
"Black box functions with uncertain inputs are used to encode knowledge in many areas of science, engineering and commerce. The challenge is to get that knowledge back out for the benefit of users. The customary approaches focus on subtle mathematics and expensive computations. This book presents SimDec, which produces interpretable graphical representations that support discussions and discovery."
Art Owen, Max H. Stein Professor of Statistics, Stanford University, USA
"The fusion of sensitivity analysis with uncertainty analysis through SimDec, as presented in this book, marks a watershed moment for business professionals and GenAI developers alike. It's a guidebook for those who dare to challenge the status quo, offering not just insights but a comprehensive toolkit for transformative decision-making. A testament to the power of interdisciplinary collaboration and open-source innovation in shaping the future of technology. Essential reading for leaders driving innovation in uncertain times."
Anferny Chen, CEO and Founder, Dataraction/DotsLive.Com, Canada
"This book articulates the SimDec method for global sensitivity analysis by combining a novel visual uncertainty analysis approach with the discriminatory capabilities of a newly created technique for calculating sensitivity indices. The real beauty of SimDec is that it can be straightforwardly applied to virtually any field of data analysis, irrespective of the mathematical sophistication of the user. I have been working with system dynamics, optimization, stochastic programming, and other analytical approaches for over three decades. One regret I now have is that there was no SimDec procedure in existence at the time to support these activities."
Gordon Huang, Canada Research Chair & Professor of Environmental Systems Engineering, University of Regina, Canada
