Emre Onal

Emre Onal

Visitor from Princeton

Columbia University

Princeton University

I am from Istanbul, Turkey and Geneva, Switzerland. I graduated summa cum laude from Princeton University in 2023, majoring in Operations Research & Financial Engineering with minors in Computer Science and Statistics & Machine Learning. Currently, I am pursuing my Master’s in Computer Science at Columbia University, which I will complete in December 2024.

I am most interested in automating complex large-scale decision-making processes using statistics and machine learning. I have a diverse range of experiences in different areas of machine learning, from computer vision to natural language processing. My past research experiences have ranged from improving our understanding of feature learning in deep classification networks to optimizing ads ranking performance for Facebook ads using graph learning models. My most recent research project explored a simple and inexpensive way to integrate Bayesian neural networks with large language models by using Gaussian stochastic weight averaging in low-rank adaptation layers, resulting in improvements in model calibration and generalization performance across several natural language processing benchmarks.

Interests
  • Generative AI / Large Language Models
  • Retrieval Augmented Generation
  • Information Retrieval
  • Natural Language Processing
  • Multimodal Tranformers
  • Bayesian Statistics
  • Genetic algorithms
Education
  • MS in Computer Science, 2024

    Columbia University

  • BSE in Operations Research & Financial Engineering; Minors in Computer Science, Statistics & Machine Learning, 2023

    Princeton University