Build Your Own Neural Network From Scratch

Samantha Donner; Ethan Geary; Peter March; Mathew Umano

Home/Archive/arXiv

arXiv · 2610.24271

Build Your Own Neural Network From Scratch

Samantha Donner·Ethan Geary·Peter March·Mathew Umano

View original source ↗Read source document ↗Export metadata ↓

Abstract

These notes are based on our reading of Michael Nielsen's online book neuralnetworksanddeeplearning.com. They represent our effort to elaborate his exposition of neural networks and to provide supporting mathematical detail. Our goal is to provide self-contained instructions for building a basic neural network in the form of pseudocode, so that an interested reader can build a network from scratch and know why it works. No prior knowledge of neural networks or of Python is assumed. Some familiarity with linear algebra, calculus, and probability is helpful but we explain the relevant mathematics in a set of appendices designed to be accessible to newcomers.

Explore related subjects

math.HO

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Samantha Donner, Ethan Geary, Peter March, Mathew Umano. 2026-10-16. Build Your Own Neural Network From Scratch. https://arxiv.org/abs/2610.24271

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

A Vector-Algebraic Reconstruction of the Medieval Arab Formula for the Shadow of a Gnomon

We present a modern reconstruction of the classical formula, first derived by medieval Arab astronomers, that describes the trajectory of the tip of a gnomon's shadow during the day as a function of latitude, solar declination, and gnomon height. Unlike the traditional derivations based on spherical trigonometry, our approach uses only elementary vector algebra and rotation matrices. The same vector framework naturally extends to other classical relations used in sundial construction.

math.HO↗

arXiv2026

The Wandering Must Continue: Pure Mathematicians as Rangers of the North

Pure mathematics is an honest barometer of a civilization's scientific capacity, maintaining an unwritten culture of rigor, intellectual honesty, and precision whose primary service is not merely theorems, but human judgment. In an era of automated theorem generation and formal verification, this essay argues that establishing a result and understanding why it holds remain distinct achievements that require deliberate institutional protection.

math.HO↗

arXiv2026

The crisis of AI-generated mathematics

In this essay, I present the case for total opposition to the use of artificial intelligence in mathematics. I offer proposals for how individuals, departments, journals, and institutions can act in concert to make sure that mathematics survives the coming crisis.

math.HO↗