# Double Descent: Understanding Linear Model Estimation of Nonidentifiable Parameters and a Model for Overfitting

arXiv: https://arxiv.org/abs/2408.13235 · HTML: https://mlsanity.com/paper/2408.13235/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 5 of its 16 references that are in the map.
- First posted on arXiv 23 Aug 2024.

## Builds on (most cited first, 5 of 5)

- [Surprises in High-Dimensional Ridgeless Least Squares Interpolation](https://mlsanity.com/paper/1903.08560/) (2019)
- [An Overview on the Estimation of Large Covariance and Precision Matrices](https://mlsanity.com/paper/1504.02995/) (2015)
- [Comparing Classes of Estimators: When does Gradient Descent Beat Ridge Regression in Linear Models?](https://mlsanity.com/paper/2108.11872/) (2021)
- [Asymptotic properties of Principal Component Analysis and shrinkage-bias adjustment under the Generalized Spiked Population model](https://mlsanity.com/paper/1607.08647/) (2016)
- [On the number of principal components in high dimensions](https://mlsanity.com/paper/1708.04981/) (2017)

## Built on by (most cited first, 1 of 1)

- [Evaluating Double Descent in Machine Learning: Insights from Tree-Based Models Applied to a Genomic Prediction Task](https://mlsanity.com/paper/2509.25216/) (2025)

## Abstract (arXiv)

We consider ordinary least squares estimation and variations on least squares estimation such as penalized (regularized) least squares and spectral shrinkage estimates for problems with p > n and associated problems with prediction of new observations. After the introduction of Section 1, Section 2 examines a number of commonly used estimators for p > n. Section 3 introduces prediction with p > n. Section 4 introduces notational changes to facilitate discussion of overfitting and Section 5 illustrates the phenomenon of double descent. We conclude with some final comments.