Data interpreter: An llm agent for data science

S Hong, Y Lin, B Liu, B Liu, B Wu, C Zhang… - arxiv preprint arxiv …, 2024‏ - arxiv.org
Large Language Model (LLM)-based agents have shown effectiveness across many
applications. However, their use in data science scenarios requiring solving long-term …

Nas-bench-suite-zero: Accelerating research on zero cost proxies

A Krishnakumar, C White, A Zela… - Advances in …, 2022‏ - proceedings.neurips.cc
Zero-cost proxies (ZC proxies) are a recent architecture performance prediction technique
aiming to significantly speed up algorithms for neural architecture search (NAS). Recent …

[PDF][PDF] Data science at the singularity

D Donoho - Harvard Data Science Review, 2024‏ - assets.pubpub.org
Something fundamental to computation-based research has really changed in the last ten
years. In certain fields, progress is simply dramatically more rapid than previously …

Meta-album: Multi-domain meta-dataset for few-shot image classification

I Ullah, D Carrión-Ojeda, S Escalera… - Advances in …, 2022‏ - proceedings.neurips.cc
Abstract We introduce Meta-Album, an image classification meta-dataset designed to
facilitate few-shot learning, transfer learning, meta-learning, among other tasks. It includes …

Dual-view molecular pre-training

J Zhu, Y **a, L Wu, S **e, W Zhou, T Qin, H Li… - Proceedings of the 29th …, 2023‏ - dl.acm.org
Molecular pre-training, which is about to learn an effective representation for molecules on
large amount of data, has attracted substantial attention in cheminformatics and …

Human behavior in image-based Road Health Inspection Systems despite the emerging AutoML

T Siriborvornratanakul - Journal of Big Data, 2022‏ - Springer
Introduction The emergence of automated machine learning or AutoML has raised an
interesting trend of no-code and low-code machine learning where most tasks in the …

Lessons learned from the NeurIPS 2021 MetaDL challenge: Backbone fine-tuning without episodic meta-learning dominates for few-shot learning image classification

A El Baz, I Ullah, E Alcobaça… - NeurIPS 2021 …, 2022‏ - proceedings.mlr.press
Although deep neural networks are capable of achieving performance superior to humans
on various tasks, they are notorious for requiring large amounts of data and computing …

Towards efficient AutoML: a pipeline synthesis approach leveraging pre-trained transformers for multimodal data

A Moharil, J Vanschoren, P Singh, D Tamburri - Machine Learning, 2024‏ - Springer
This paper introduces an Automated Machine Learning (AutoML) framework specifically
designed to efficiently synthesize end-to-end multimodal machine learning pipelines …

NeurIPS'22 cross-domain MetaDL challenge: Results and lessons learned

D Carrión-Ojeda, M Alam, S Escalera… - NeurIPS 2022 …, 2023‏ - proceedings.mlr.press
Deep neural networks have demonstrated the ability to outperform humans in multiple tasks,
but they often require substantial amounts of data and computational resources. These …

Visda-2021 competition: Universal domain adaptation to improve performance on out-of-distribution data

D Bashkirova, D Hendrycks, D Kim… - NeurIPS 2021 …, 2022‏ - proceedings.mlr.press
Progress in machine learning is typically measured by training and testing a model on
samples drawn from the same distribution, ie the same domain. This over-estimates future …