Interests

My primary research interests are Causality and Machine Learning (ML), with an emphasis on using modern ML methods to solve causal problems. More specific areas of interest are as follows.

Causality

• Treatment Effect Estimation
• Structure Learning / Causal Discovery
• Observational Data
• Cross-sectional and Panel Data
• Treatment Recommendation / Policy Learning

Machine Learning

• Hyperparameter Tuning and Model Selection
• Performance Evaluation
• Data and Covariate Shifts
• OOD Generalisation

Research Software Engineering

I am also interested in Software Engineering for Research purposes, which nicely fits into my software development background. Some of the topics and themes that particularly interest me are:

• Backend Development
• Benchmarking
• Reproducibility
• Performance Engineering

Publications

• D. Machlanski, S. Samothrakis, and P. Clarke. (2024). Robustness of Algorithms for Causal Structure Learning to Hyperparameter Choice. Proceedings of the Third Conference on Causal Learning and Reasoning, in Proceedings of Machine Learning Research 236:703-739. [paper] [code]
• D. Machlanski, S. Samothrakis, and P. Clarke. (2024). Undersmoothing Causal Estimators With Generative Trees. IEEE Access, vol. 12, pp. 38562-38574. [paper] [code]

Working Papers

• D. Machlanski, S. Samothrakis, and P. Clarke, ‘Hyperparameter Tuning and Model Evaluation in Causal Effect Estimation’. arXiv, Mar. 02, 2023. doi: 10.48550/arXiv.2303.01412. [paper] [code]

Presentations

• The Importance of Hyperparameter Tuning in Causal Effect Estimation, Causal Data Science Meeting 2022 [slides].
• Causal Discovery for Treatment Effect Estimation from Observational Data, MiSoC Annual Workshop 2021.

Other Code Repositories

• CATE Benchmark [github]

Find me on

• Google Scholar
• ResearchGate
• GitHub
• LinkedIn