# To go deep or wide in learning?

arXiv: https://arxiv.org/abs/1402.5634 · HTML: https://mlsanity.com/paper/1402.5634/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 Feb 2014.

## Built on by (most cited first, 4 of 4)

- [Recommender systems based on graph embedding techniques: A comprehensive review](https://mlsanity.com/paper/2109.09587/) (2021)
- [Physics-informed Neural Networks approach to solve the Blasius function](https://mlsanity.com/paper/2301.00106/) (2022)
- [Neural Generalization of Multiple Kernel Learning](https://mlsanity.com/paper/2102.13337/) (2021)
- [Wide Learning: Learning to Reach Evidence](https://mlsanity.com/paper/2608.29608/) (2026)

## Abstract (arXiv)

To achieve acceptable performance for AI tasks, one can either use sophisticated feature extraction methods as the first layer in a two-layered supervised learning model, or learn the features directly using a deep (multi-layered) model. While the first approach is very problem-specific, the second approach has computational overheads in learning multiple layers and fine-tuning of the model. In this paper, we propose an approach called wide learning based on arc-cosine kernels, that learns a single layer of infinite width. We propose exact and inexact learning strategies for wide learning and show that wide learning with single layer outperforms single layer as well as deep architectures of finite width for some benchmark datasets.