---
date: '2024-01-11'
description: neural networks learning hierarchical representations by detecting patterns from simple edges to complex features across layers.
id: deep learning
modified: 2026-06-05 15:08:29 GMT-04:00
tags:
  - ml
  - framework
title: deep learning
created: '2024-01-11'
published: '2024-01-11'
pageLayout: default
slug: thoughts/deep-learning
permalink: https://aarnphm.xyz/thoughts/deep-learning.md
generator:
  quartz: v4.6.0
  hostedProvider: Cloudflare
  baseUrl: aarnphm.xyz
full: https://aarnphm.xyz/llms-full.txt
---
See also: <ref slug="books#2024"> ([pdf](https://fleuret.org/public/lbdl.pdf) or [lectures](https://fleuret.org/dlc/)), [annotated history of deep learning](https://people.idsia.ch/~juergen/deep-learning-history.html), this [lecture series at CMU](https://dlsyscourse.org/lectures/)

> \[!eli5\] deep learning
>
> Imagine you’re learning to recognize dogs. At first, your parents point to different dogs and say “that’s a dog!” After seeing lots of dogs, you start noticing patterns - they have four legs, fur, tails, and make barking sounds. Now you can spot dogs on your own!
>
> Deep learning works kind of like that, but for computers. The computer looks at tons of pictures (like thousands and thousands), and slowly figures out what makes a dog look like a dog. It starts with simple things like edges and shapes, then builds up to more complicated stuff like spotting ears, tails, and finally whole dogs.

Deep learning is a superset of [[thoughts/Machine learning]]. [[thoughts/university/twenty-four-twenty-five/sfwr-4ml3|Supervised]] and un-supervised refers to the training objective within machine learning, rather than
the goal of different algorithms.