# Ayman Boustati

## Profile:

I am a fourth year PhD student in the Mathematic for Real-World Systems CDT under the supervision of Dr Theo Damoulas. My main area of research is in Probabilistic Machine Learning. I work on developing novel modelling and inference, with a particular focus on Gaussian process models.

## Publications:

• Boustati, A., Vakili, S., Hensman, J., & John, ST (2020). Amortized variance reduction for doubly stochastic objectives. In Proceedings of the 36th Conference on Uncertainty in Artificial Intelligence (UAI), PMLR volume 124. https://arxiv.org/abs/2003.04125.
• Richter, L., Boustati, A.*, Nüksen, N., Ruiz, F. J., Akyildiz, Ö. D. (2020). VarGrad: A Low Variance Gradient Estimator for Variational Inference. Under Review.
• Boustati, A., Akyildiz, Ö. D., Damoulas, T., & Johansen, A. (2020). Generalized Bayesian Filtering via Sequential Monte Carlo. arXiv preprint arXiv:2002.09998. Under Review. https://arxiv.org/abs/2003.04125.
• Boustati, A., Damoulas, T., & Savage, R. S. (2019). Non-linear Multitask Learning with Deep Gaussian Processes. arXiv preprint arXiv:1905.12407. Under Review. https://arxiv.org/abs/1905.12407.

* Joint first author.

## Background:

• MSc in Mathematics for Real-World Systems from the University of Warwick 2015-2016
• Master of Mathematics, Operational Research, Statistics and Economics from the University of Warwick 2011- 2015

## Research Interests:

• Gaussian Processes
$\mbox{a.boustati@warwick.ac.uk}$