TabM: Advancing Tabular Deep Learning with Parameter-Efficient Ensembling

人工智能 计算机科学
作者
Yury Gorishniy,Akim Kotelnikov,Artem Babenko
出处
期刊:Cornell University - arXiv [Cornell University]
被引量:178
标识
DOI:10.48550/arxiv.2410.24210
摘要

Deep learning architectures for supervised learning on tabular data range from simple multilayer perceptrons (MLP) to sophisticated Transformers and retrieval-augmented methods. This study highlights a major, yet so far overlooked opportunity for designing substantially better MLP-based tabular architectures. Namely, our new model TabM relies on efficient ensembling, where one TabM efficiently imitates an ensemble of MLPs and produces multiple predictions per object. Compared to a traditional deep ensemble, in TabM, the underlying implicit MLPs are trained simultaneously, and (by default) share most of their parameters, which results in significantly better performance and efficiency. Using TabM as a new baseline, we perform a large-scale evaluation of tabular DL architectures on public benchmarks in terms of both task performance and efficiency, which renders the landscape of tabular DL in a new light. Generally, we show that MLPs, including TabM, form a line of stronger and more practical models compared to attention- and retrieval-based architectures. In particular, we find that TabM demonstrates the best performance among tabular DL models. Then, we conduct an empirical analysis on the ensemble-like nature of TabM. We observe that the multiple predictions of TabM are weak individually, but powerful collectively. Overall, our work brings an impactful technique to tabular DL and advances the performance-efficiency trade-off with TabM -- a simple and powerful baseline for researchers and practitioners.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
小马甲的应助被科研通管家采纳,获得10
刚刚
小蘑菇的应助被科研通管家采纳,获得10
1秒前
Jun的应助被科研通管家采纳,获得10
1秒前
隐形曼青的应助被科研通管家采纳,获得10
1秒前
369ninja的应助被瑾风阳采纳,获得10
1秒前
1秒前
zengyp23完成签到,获得积分10
1秒前
脑洞疼的应助被科研通管家采纳,获得10
1秒前
1秒前
1秒前
1秒前
彭于晏的应助被科研通管家采纳,获得10
1秒前
赘婿的应助被科研通管家采纳,获得10
2秒前
Ray的应助被科研通管家采纳,获得10
2秒前
2秒前
2秒前
orixero的应助被st5114采纳,获得10
2秒前
隐形曼青的应助被科研通管家采纳,获得10
2秒前
Migludia完成签到,获得积分10
2秒前
Jun的应助被科研通管家采纳,获得10
2秒前
2秒前
英俊的铭的应助被科研通管家采纳,获得10
2秒前
阿智发布了新的文献求助10
2秒前
2秒前
小蘑菇的应助被科研通管家采纳,获得10
2秒前
上官若男的应助被科研通管家采纳,获得30
3秒前
英俊的铭的应助被科研通管家采纳,获得10
3秒前
lx的应助被科研通管家采纳,获得10
3秒前
领导范儿的应助被科研通管家采纳,获得10
3秒前
斯文败类的应助被科研通管家采纳,获得10
3秒前
共享精神的应助被科研通管家采纳,获得10
3秒前
3秒前
4秒前
灵萧莫言完成签到,获得积分10
5秒前
晚意发布了新的文献求助10
5秒前
5秒前
6秒前
青青完成签到,获得积分10
6秒前
Echo发布了新的文献求助10
6秒前
6秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
Grapes 1000
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
John McDowells Bildungstheorie 800
八旗制度の研究 600
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8178375
求助须知:如何正确求助?哪些是违规求助? 9569182
关于积分的说明 21802781
捐赠科研通 7773247
什么是DOI,文献DOI怎么找? 3432799
关于科研通互助平台的介绍 2556793
邀请新用户注册赠送积分活动 2478346