When Do Neural Nets Outperform Boosted Trees on Tabular Data? Tabular data is one of the most commonly used types of data in machine learning. Despite recent advances in neural nets (NNs) for tabu...

Distributionally Robust Classification on a Data Budget Real world uses of deep learning require predictable model behavior under distribution shifts. Models such as CLIP show emergent natural dis...

Image classification and object detection in agriculture One important real-world domain for applying deep learning is agriculture. Over the last two years, under the auspices of AIIRA, a collabor...

ArcheType This post describes ArcheType, a new way of doing column type annotation using large language models under the hood, which I developed under the supervision of Prof. Freire and Prof. Heg...