Metamodeling atomic models in discrete event system specification (devs) formalism using multivariate adaptive regression splines (mars)

Bozağaç, Cumhur Doruk
Computer simulations are widely used for design optimization purposes. The problem becomes challenging when design variables are high dimensional and when the simulation is computationally expensive. In this work we propose a methodology for metamodeling of dynamic simulation models via Multivariate Adaptive Regres sion Splines (MARS). To handle incomplete output processes, where the simulation model does not produce an output in some steps due to missing inputs, we have devised a two-level metamodeling scheme. The methodology is demonstrated on a dynamic radar simulation model. The prediction performance of the resulting meta model is tested with four different sampling techniques (i.e., experimental designs) and 16 sample sizes. We also investigate the effect of alternative coordinate system representations on the metamodeling performance. The results suggest that MARS is an effective method for metamodeling dynamic simulations, particularly, when expert judgment is not readily available. Results also show that there are interactions between the coordinate system representations and sampling techniques, and some sampling-representation-size combinations are very promising in the solution to this type of problem. The technique is then applied to develop proxy models (PMs) of atomic models in Discrete Event System Specification (DEVS) simulations to replace an atomic model with a PM that uses the fitted metamodel. The mechanisms required for replacing an atomic model and integrating the PM into a simulation are described in detail. The methodology is tested by replacing a radar atomic model in a military engagement simulation with the PM and the method’s benefits and chal- lenges are discussed thoroughly. As a final step of this research, the response surfaces of the original simulation and the PM integrated simulation is analyzed. Both simulations are used in a Particle Swarm Optimization based optimization procedure and the results are compared. Results obtained from this particular case suggest that metamodeling of computationally intensive atomic models is feasible, and even with relatively small number of observations, we can apply metamodeling to sub-components of DEVS simulation models successfully.


Teran-Somohano, Alejandro; Dayibas, Orcun; Yılmaz, Levent; Smith, Alice (2014-12-10)
Goal-directed reproducible experimentation with simulation models is still a significant challenge. The underutilization of design of experiments, limited transparency in the collection and analysis of results, and ad-hoc adaptation of experiments as learning takes place continue to hamper reproducibility and hence cause a credibility gap. In this study, we propose a strategy that leverages the synergies between model-driven engineering, intelligent agent technology, and variability modeling to support the ...
Parametric Urban Design Thinking: Shared Patterns in Design by Algorithm and Design by Drawing
Çalışkan, Olgu; Ongun, Gokhan (2021-12-01)
The paper suggests a focused examination of the processes of drafting-based design and parametric design in urbanism. It discusses how spatial design's settled cognition would differ by using algorithmic systems through the altered relationships between the basic operations in design. To reveal the commonalities and distinctions between the two design methods, the authors present the detailed documentation of the workshop series, which experimented with both techniques within similar design contexts. By the...
Performance-based parametric design explorations: A method for generating appropriate building components
Ercan, Burak; Elias Özkan, Soofia Tahira (2015-05-01)
Performance-based parametric design explorations depend on formulating custom-designed workflows that require reading, writing, interpreting and manipulating databases, as part of the design process. The possibilities of customization and parameterization offered by the user-friendly interfaces of advanced building-performance simulation software and digital design tools have now enabled architects to carry out performance-based design explorations without the help of simulation experts. This paper presents...
Extending Correlation Filter-Based Visual Tracking by Tree-Structured Ensemble and Spatial Windowing
Gundogdu, Erhan; Özkan, Huseyin; Alatan, Abdullah Aydın (Institute of Electrical and Electronics Engineers (IEEE), 2017-11-01)
Correlation filters have been successfully used in visual tracking due to their modeling power and computational efficiency. However, the state-of-the-art correlation filter-based (CFB) tracking algorithms tend to quickly discard the previous poses of the target, since they consider only a single filter in their models. On the contrary, our approach is to register multiple CFB trackers for previous poses and exploit the registered knowledge when an appearance change occurs. To this end, we propose a novel t...
Parametric Design in Urbanism: A Critical Reflection
Çalışkan, Olgu (2017-01-01)
Parametric modelling run by the explicitly defined algorithms generating synchronically auditable dynamic forms and patterns, has become a prominent method especially in architecture. Though the use of parametric models has got wider in urban design, the critical reflection on the actual and possible application of the method in urbanism has fallen limited so far. The paper tends to relate parametric design with the contemporary understanding of urbanism with regards to the idea of design control in the the...
Citation Formats
C. D. Bozağaç, “Metamodeling atomic models in discrete event system specification (devs) formalism using multivariate adaptive regression splines (mars),” Ph.D. - Doctoral Program, Middle East Technical University, 2014.