email@example.com / (213) 740-5910/ AHF B56G
2016 Ph.D., Spatial Statistics; minor, Geological Sciences, Stanford University
2012 M.S., Geostatistics, Stanford University
2010 B.S., Electrical and Electronic Engineering; B.S., Petroleum and Natural Gas Engineering, Middle East Technical University, Ankara, Turkey
Orhun Aydin, Ph.D., is a Lecturer with the Spatial Sciences Institute at the University of Southern California Dornsife College of Letters, Arts and Sciences. He is also a researcher in Esri's Spatial Statistics team.
His research interests pertain to forecasting and understanding Earth systems via spatial and spatio-temporal statistical models. His theoretical work involves developing general mathematical frameworks for modelling, mining and representing spatial and spatio-temporal data. In particular, he researches graph-based methods to represent spatial data in machine learning applications. His applied research combines computer vision, machine learning, spatial statistics and Earth process models to develop domain-knowledge driven machine learning methods for modelling Earth processes under uncertainty and sparse data. As a lecturer at USC, he is teaching courses on spatio-temporal statistics and spatial statistics.
He has been publishing journal papers on applied spatial machine learning in marine ecology, tectonophysics and extreme weather events. His published theoretical work spans level-sets, marked point process models, spatial graph models and consensus-based regionalization. His research on consensus-based regionalization won an award in 2018 ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems in Seattle, WA.
Aydin, O., Janikas, M. V., Assunção, R., & Lee, T. H. (2018, November). SKATER-CON: Unsupervised Regionalization via Stochastic Tree Partitioning within a Consensus Framework Using Random Spanning Trees. In Proceedings of the 2nd ACM SIGSPATIAL International Workshop on AI for Geographic Knowledge Discovery (pp. 33-42). ACM.
Aydin, O., Butler, K., & Wright, D. (2018, April). Integrated GIS/Machine-Learning Workflows for Modeling Spatiotemporal Variations in Potential Seagrass Habitats within a Changing Climate. In EGU General Assembly Conference Abstracts (Vol. 20, p. 10081).
Aydin, O., & Caers, J. K. (2017). Quantifying structural uncertainty on fault networks using a marked point process within a Bayesian framework. Tectonophysics, 712, 101-124.
Aydin, O., & Caers, J. K. (2014, October). Exploring structural uncertainty using a flow proxy in the depositional domain. In SPE Annual Technical Conference and Exhibition. Society of Petroleum Engineers.
Aydin, O., & Caers, J. (2013). Image transforms for determining fit-for-purpose complexity of geostatistical models in flow modeling. Computational Geosciences, 17(2), 417-429.