|
Automated machine learning
Data cleaning, feature selection, pipeline design, and hyperparameter tuning remain some of the most challenging (and time-consuming) tasks in data science. My work on automated machine learning (AutoML) seeks to automate these tasks to enable widespread use of machine learning by non-experts.
Talks
Software
Papers
Resource-Constrained Neural Architecture Search on Tabular Datasets
C. Yang, G. Bender, H. Liu, P. Kindermans, M. Udell, Y. Lu, Q. Le, and D. Huang
NeurIPS, 2022
[arxiv][url][bib]
How Low Can We Go: Trading Memory for Error in Low-Precision Training
C. Yang, Z. Wu, J. Chee, C. D. Sa, and M. Udell
International Conference on Learning Representations (ICLR), 2022
[arxiv][url][bib]
Can we globally optimize cross-validation loss? Quasiconvexity in ridge regression
W. T. Stephenson, Z. Frangella, M. Udell, and T. Broderick
Advances in Neural Information Processing Systems (NeurIPS), 2021
[arxiv][bib]
Privileged Zero-Shot AutoML
N. Singh, B. Kates, J. Mentch, A. Kharkar, M. Udell, and I. Drori
2021
[arxiv][bib]
Efficient AutoML Pipeline Search with Matrix and Tensor Factorization
C. Yang, J. Fan, Z. Wu, and M. Udell
2020
[arxiv][url][bib]
Real-time AutoML
I. Drori, L. Liu, Q. Ma, J. Deykin, B. Kates, and M. Udell
2020
[bib]
AutoML Pipeline Selection: Efficiently Navigating the Combinatorial Space
C. Yang, J. Fan, Z. Wu, and M. Udell
ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), 2020
[arxiv][url][bib]
AutoML using Metadata Language Embeddings
I. Drori, L. Liu, S. Koorathota, N. Yi, J. Li, A. Moretti, J. Freire, and M. Udell
NeurIPS Workshop on Meta-Learning, 2019
[arxiv][pdf][bib]
OBOE: Collaborative Filtering for AutoML Model Selection
C. Yang, Y. Akimoto, D. Kim, and M. Udell
ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), 2019
Oral presentation
[arxiv][pdf][url][bib]
OBOE: Collaborative Filtering for AutoML Initialization (workshop version)
C. Yang, Y. Akimoto, D. Kim, and M. Udell
NeurIPS Workshop on Automated Machine Learning, 2018
[arxiv][pdf][url][bib]
|