Amirarsalan Rajabi’s Homepage
Bio
I am a Senior Machine Learning Engineer at Integral Ad Science, where I build and deploy large-scale LLM and vision-language systems — spanning distributed inference infrastructure, automated labeling pipelines, and production ML frameworks. I earned my Ph.D. from the University of Central Florida in December 2022, where I worked as a graduate research assistant at CASL.
Selected Publications
Fair Bilevel Neural Network (FairBiNN): On Balancing Fairness and Accuracy via Stackelberg Equilibrium
NeurIPS 2024
Fairness and accuracy are treated as two players in a Stackelberg game, each owning its own layers of one network. The bilevel solution is provably no worse than the usual Lagrangian (regularization) approach, and in practice it reaches a better accuracy–fairness trade-off.
Through a Fair Looking-Glass: Mitigating Bias in Image Datasets
HCII 2023
A U-Net learns small edits to each image that reduce the statistical dependence (HSIC) between the target attribute (Attractive) and gender. This de-biases the dataset itself, so classifiers trained on it are fairer.
TabFairGAN: Fair Tabular Data Generation with Generative Adversarial Networks
MAKE 2022
A Wasserstein GAN trained in two phases: first to reproduce the joint distribution of a tabular dataset, then with a fairness term added to the generator loss. The result is synthetic data with demographic parity nearly removed and utility largely kept.
Other Publications
- Distance Correlation GAN: Fair Tabular Data Generation with Generative Adversarial Networks
HCII 2023 - Exploring the disparity of influence between users in the discussion of Brexit on Twitter
J. Comput. Soc. Sci. 2021 - Review on learning and extracting graph features for link prediction
MAKE 2020 - Polarization in social media assists influencers to become more influential: analysis and two inoculation strategies
Nature Scientific Reports 2019