My research interests are broad and interdisciplinary. Aside from my primary focus areas (i.e. conflict, IR, methods), I have been involved in projects ranging from environmental sustainability to extremist online chatter. Indicatively, some of my current endeavors involve research on the detection of underlying bias in multicluster data, the development of DNN survival analysis algorithms, and the use of AI in assessing freshwater quality in New Zealand.
Published Research
Pullar-Strecker, Zac, Xinglong Chang, Liam Brydon, Ioannis Ziogas, Katharina Dost, and Joerg Wicker. (Forthcoming). “Memento: Facilitating Effortless, Efficient, and Reliable ML Experiments.” European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases.
Dost, Katharina, Hamish Duncanson, Ioannis Ziogas, Patricia Riddle, and Jörg Wicker. 2022. "Divide and Imitate: Multi-Cluster Identification and Mitigation of Selection Bias." Advances in Knowledge Discovery and Data Science (May 2022): 149-160. Link.
Willis, Derek, Merivaki, Thessalia, and Ioannis Ziogas. 2021. "Election Data Transparency: Obtaining Precinct-Level Election Returns." Public Integrity 24(2): 162-178. Link.
Ongoing Projects
My research is currently focused on the following topics:
Predicting ovine methane emissions based on rumen microbiome characteristics (with Sandeep Gupta and Joerg Wicker).
Deep neural network architecture for nonparametric survival analysis under the presence of unconditional frailty.
Bias detection in non-Gaussian, categorical, and multi-cluster data (with Kat Dost & Joerg Wicker).
Evaluation of protosampling algorithms in conflict data (with Chrysafis Vogiatzis).
Development of online extremism threat assessment tools (with Colleen Sinclair).
State emergence and interstate conflict.
AI and freshwater modelling in New Zealand (multi-member, interdisciplinary team).
Data
I have collected and compiled data on State Emergence Processes (SEP) from 1800-2010. I expect the newest version (SEPv4.1) will be ready for release in mid-2023.