Position: Bayesian Deep Learning is Needed in the Age of Large-Scale AI

Theodore Papamarkou; Maria Skoularidou; Konstantina Palla; Laurence Aitchison; Julyan Arbel; David Dunson; Maurizio Filippone; Vincent Fortuin; Philipp Hennig; José Miguel Hernández-Lobato; Aliaksandr Hubin; Alexander Immer; Theofanis Karaletsos; Mohammad Emtiyaz Khan; Agustinus Kristiadi; Yingzhen Li; Stephan Mandt; Christopher Nemeth; Michael A. Osborne; Tim G.J. Rudner; David Rügamer; Yee Whye Teh; Max Welling; Andrew Gordon Wilson; Ruqi Zhang

Position: Bayesian Deep Learning is Needed in the Age of Large-Scale AI

Theodore Papamarkou, Maria Skoularidou, Konstantina Palla, Laurence Aitchison, Julyan Arbel, David Dunson, Maurizio Filippone, Vincent Fortuin, Philipp Hennig, José Miguel Hernández-Lobato, Aliaksandr Hubin, Alexander Immer, Theofanis Karaletsos, Mohammad Emtiyaz Khan, Agustinus Kristiadi, Yingzhen Li, Stephan Mandt, Christopher Nemeth, Michael A. Osborne, Tim G.J. RudnerDavid Rügamer, Yee Whye Teh, Max Welling, Andrew Gordon Wilson, Ruqi Zhang

Computer Science

Research output: Contribution to journal › Conference article › peer-review

Abstract

In the current landscape of deep learning research, there is a predominant emphasis on achieving high predictive accuracy in supervised tasks involving large image and language datasets. However, a broader perspective reveals a multitude of overlooked metrics, tasks, and data types, such as uncertainty, active and continual learning, and scientific data, that demand attention. Bayesian deep learning (BDL) constitutes a promising avenue, offering advantages across these diverse settings. This paper posits that BDL can elevate the capabilities of deep learning. It revisits the strengths of BDL, acknowledges existing challenges, and highlights some exciting research avenues aimed at addressing these obstacles. Looking ahead, the discussion focuses on possible ways to combine large-scale foundation models with BDL to unlock their full potential.

Original language	English (US)
Pages (from-to)	39556-39586
Number of pages	31
Journal	Proceedings of Machine Learning Research
Volume	235
State	Published - 2024
Event	41st International Conference on Machine Learning, ICML 2024 - Vienna, Austria Duration: Jul 21 2024 → Jul 27 2024

ASJC Scopus subject areas

Artificial Intelligence
Software
Control and Systems Engineering
Statistics and Probability

Cite this

Papamarkou, T., Skoularidou, M., Palla, K., Aitchison, L., Arbel, J., Dunson, D., Filippone, M., Fortuin, V., Hennig, P., Hernández-Lobato, J. M., Hubin, A., Immer, A., Karaletsos, T., Khan, M. E., Kristiadi, A., Li, Y., Mandt, S., Nemeth, C., Osborne, M. A., ... Zhang, R. (2024). Position: Bayesian Deep Learning is Needed in the Age of Large-Scale AI. Proceedings of Machine Learning Research, 235, 39556-39586.

Papamarkou, T, Skoularidou, M, Palla, K, Aitchison, L, Arbel, J, Dunson, D, Filippone, M, Fortuin, V, Hennig, P, Hernández-Lobato, JM, Hubin, A, Immer, A, Karaletsos, T, Khan, ME, Kristiadi, A, Li, Y, Mandt, S, Nemeth, C, Osborne, MA, Rudner, TGJ, Rügamer, D, Teh, YW, Welling, M, Wilson, AG & Zhang, R 2024, 'Position: Bayesian Deep Learning is Needed in the Age of Large-Scale AI', Proceedings of Machine Learning Research, vol. 235, pp. 39556-39586.

@article{9b42a217bd2d4a74af1987c750ddd830,

title = "Position: Bayesian Deep Learning is Needed in the Age of Large-Scale AI",

abstract = "In the current landscape of deep learning research, there is a predominant emphasis on achieving high predictive accuracy in supervised tasks involving large image and language datasets. However, a broader perspective reveals a multitude of overlooked metrics, tasks, and data types, such as uncertainty, active and continual learning, and scientific data, that demand attention. Bayesian deep learning (BDL) constitutes a promising avenue, offering advantages across these diverse settings. This paper posits that BDL can elevate the capabilities of deep learning. It revisits the strengths of BDL, acknowledges existing challenges, and highlights some exciting research avenues aimed at addressing these obstacles. Looking ahead, the discussion focuses on possible ways to combine large-scale foundation models with BDL to unlock their full potential.",

author = "Theodore Papamarkou and Maria Skoularidou and Konstantina Palla and Laurence Aitchison and Julyan Arbel and David Dunson and Maurizio Filippone and Vincent Fortuin and Philipp Hennig and Hern{\'a}ndez-Lobato, {Jos{\'e} Miguel} and Aliaksandr Hubin and Alexander Immer and Theofanis Karaletsos and Khan, {Mohammad Emtiyaz} and Agustinus Kristiadi and Yingzhen Li and Stephan Mandt and Christopher Nemeth and Osborne, {Michael A.} and Rudner, {Tim G.J.} and David R{\"u}gamer and Teh, {Yee Whye} and Max Welling and Wilson, {Andrew Gordon} and Ruqi Zhang",

year = "2024",

language = "English (US)",

volume = "235",

pages = "39556--39586",

journal = "Proceedings of Machine Learning Research",

issn = "2640-3498",

publisher = "ML Research Press",

}

TY - JOUR

T1 - Position

T2 - 41st International Conference on Machine Learning, ICML 2024

AU - Papamarkou, Theodore

AU - Skoularidou, Maria

AU - Palla, Konstantina

AU - Aitchison, Laurence

AU - Arbel, Julyan

AU - Dunson, David

AU - Filippone, Maurizio

AU - Fortuin, Vincent

AU - Hennig, Philipp

AU - Hernández-Lobato, José Miguel

AU - Hubin, Aliaksandr

AU - Immer, Alexander

AU - Karaletsos, Theofanis

AU - Khan, Mohammad Emtiyaz

AU - Kristiadi, Agustinus

AU - Li, Yingzhen

AU - Mandt, Stephan

AU - Nemeth, Christopher

AU - Osborne, Michael A.

AU - Rudner, Tim G.J.

AU - Rügamer, David

AU - Teh, Yee Whye

AU - Welling, Max

AU - Wilson, Andrew Gordon

AU - Zhang, Ruqi

PY - 2024

Y1 - 2024

N2 - In the current landscape of deep learning research, there is a predominant emphasis on achieving high predictive accuracy in supervised tasks involving large image and language datasets. However, a broader perspective reveals a multitude of overlooked metrics, tasks, and data types, such as uncertainty, active and continual learning, and scientific data, that demand attention. Bayesian deep learning (BDL) constitutes a promising avenue, offering advantages across these diverse settings. This paper posits that BDL can elevate the capabilities of deep learning. It revisits the strengths of BDL, acknowledges existing challenges, and highlights some exciting research avenues aimed at addressing these obstacles. Looking ahead, the discussion focuses on possible ways to combine large-scale foundation models with BDL to unlock their full potential.

AB - In the current landscape of deep learning research, there is a predominant emphasis on achieving high predictive accuracy in supervised tasks involving large image and language datasets. However, a broader perspective reveals a multitude of overlooked metrics, tasks, and data types, such as uncertainty, active and continual learning, and scientific data, that demand attention. Bayesian deep learning (BDL) constitutes a promising avenue, offering advantages across these diverse settings. This paper posits that BDL can elevate the capabilities of deep learning. It revisits the strengths of BDL, acknowledges existing challenges, and highlights some exciting research avenues aimed at addressing these obstacles. Looking ahead, the discussion focuses on possible ways to combine large-scale foundation models with BDL to unlock their full potential.

UR - http://www.scopus.com/inward/record.url?scp=85203823735&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85203823735&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:85203823735

SN - 2640-3498

VL - 235

SP - 39556

EP - 39586

JO - Proceedings of Machine Learning Research

JF - Proceedings of Machine Learning Research

Y2 - 21 July 2024 through 27 July 2024

ER -

Position: Bayesian Deep Learning is Needed in the Age of Large-Scale AI

Abstract

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this