Published August 30, 2023 | Version v1
Conference paper Open

[neuralnet]: A Pure Data External for the Creation of Neural Networks Written in Pure C

Description

https://aimc2023.pubpub.org/pub/3j3fx7y1

[neuralnet] is a Pure Data external object that allows the creation of densely connected Neural Networks of any structure. This object is written entirely in C, without any dependencies. The code is a translation from Python code to C. The Python code is taken from the book Neural Networks from Scratch in Python. The aim is to create a Pd object that can be easily compiled for all OSes, so it can be easily used by Pd users of any level of expertise. This is a lightweight object that is easy to set up, train, validate, and use for predictions. Its covers three modes of operation, including classification, regression, and binary logistic regression, and it has many settable parameters, so users can customise it to their own needs. This is an open-source project.

Files

drymonitis_12_aimc2023.pdf

Files (302.9 kB)

Name Size Download all
md5:fefd096623383f8f898a1e2f185d44d5
302.9 kB Preview Download

Additional details

Related works

Is part of
978-0-9957862-9-5 (ISBN)