mlsanityData to 14 Oct 2026

A Survey of Learning Curves with Bad Behavior: or How More Data Need Not Lead to Better Performance

arXiv 2211.14061 ↗First posted 25 Nov 2022cs.LG stat.ME stat.MLcited by 2

Not on the map. Why?

Abstract (arXiv)

Plotting a learner's generalization performance against the training set size results in a so-called learning curve. This tool, providing insight in the behavior of the learner, is also practically valuable for model selection, predicting the effect of more training data, and reducing the computational complexity of training. We set out to make the (ideal) learning curve concept precise and briefly discuss the aforementioned usages of such curves. The larger part of this survey's focus, however, is on learning curves that show that more data does not necessarily leads to better generalization performance. A result that seems surprising to many researchers in the field of artificial intelligence. We point out the significance of these findings and conclude our survey with an overview and discussion of open problems in this area that warrant further theoretical and empirical investigation.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 24 of its 92 references that are in the map. First posted on arXiv 25 Nov 2022.

Builds on

24 of 92 references.

References · 24

  1. 2018Reconciling modern machine learning practice and the bias-variance trade-offcited by 927
  2. 2019Deep Double Descent: Where Bigger Models and More Data Hurtcited by 672
  3. 2017Deep Learning Scaling is Predictable, Empiricallycited by 582
  4. 2019Surprises in High-Dimensional Ridgeless Least Squares Interpolationcited by 513
  5. 2017High-dimensional dynamics of generalization error in neural networkscited by 241
  6. 2018Characterizing and Avoiding Negative Transfercited by 229
  7. 2020Optimal Regularization Can Mitigate Double Descentcited by 109
  8. 2014Inconsistency of Bayesian Inference for Misspecified Linear Models, and a Proposal for Repairing Itcited by 95
  9. 2021The Shape of Learning Curves: a Reviewcited by 70
  10. 2018A jamming transition from under- to over-parametrization affects loss landscape and generalizationcited by 67
  11. 2020Triple descent and the two kinds of overfitting: Where & why do they appear?cited by 65
  12. 2020A Brief Prehistory of Double Descentcited by 44
12 more
  1. 2019More Data Can Hurt for Linear Regression: Sample-wise Double Descentcited by 37
  2. 2022Learning Curves for Decision Making in Supervised Machine Learning: A Surveycited by 27
  3. 2020Learning Curves for Analysis of Deep Networkscited by 20
  4. 2020A critical analysis of metrics used for measuring progress in artificial intelligencecited by 14
  5. 2015Introducing Geometry in Active Learning for Image Segmentationcited by 12
  6. 2019Minimizers of the Empirical Risk and Risk Monotonicitycited by 11
  7. 2020Risk-Monotonicity in Statistical Learningcited by 9
  8. 2022Monotone Learningcited by 9
  9. 2019Making Learners (More) Monotonecited by 6
  10. 2016The Peaking Phenomenon in Semi-supervised Learningcited by 5
  11. 2021A universally consistent learning rule with a universally monotone errorcited by 4
  12. 2017Supervised Classification: Quite a Brief Overviewcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026TuneAhead: Predicting Fine-tuning Performance Before Full Training Beginscited by 1
  2. 2026Data-Efficient Deep Learning: Empirical Guidelines for Training Set Size Estimation in Inertial Sensor Classificationcited by 0