Learn The Basic Ideas of Neural Networks in 7 Pages

Published on 19 Jan 2018 ⸱ 3 min read ⸱ 17 views

Artificial intelligence is all the craze lately. Actually, it has been a matter of interest for over 50 years! Only intellectuals from domains of neurosciences, physics, computer engineering, engineering, psychology etc. have oscillated their interest in the field for various reasons. One major reason interest has only recently peaked is because the software needs large amounts of processing power. Processing power that is today available widely and cheaply. Even research, bare bone software and mountains of tagged data is now open source and ready to go out of the box. Anyone with interest in the magic of artificial intelligence can set up something crude and get going. Artificial Intelligence is effectively democratised.

What really is artificial intelligence though? I've touched upon a bunch of definitions of artificial intelligence in this post of mine. The field is fundamentally about artificially recreating the connections between nodes in the human brain. These nodes and their interaction via connections is what is called a neural network. It is what governs everything we do as humans.

I've tried my hand at learning artificial intelligence and neural networks. Trust me, it's not so easy to understand from the tutorials out there. Learning how to run a basic AI program may be easy if you follow a tutorial. But if you want a fundamental understanding of the why's, what's and how's, I've not come across anything remotely simple - explained in plain English. Until now.

I chanced upon this 7 page PDF on HackerNews explaining the Basic Ideas of Neural Networks. Unlike most other white papers that claim to give you a basic understanding, this white paper stays true to its title. It is literally the simplest explanation I've read of neural networks and the ideas around it. Four pages in and I felt like I've been learning neural networks for a couple of years. I highly recommend reading this white paper even if you're not interested in neural networks and AI. It is pure joy to see such a complex concept explained in such plain language. It's delightful and filled with knowledge.

Click the image below to download the white paper. Here's a whole list of other white papers released by Professor Bernard Widrow - Co-author of this paper

Onward.

Note: All blogs posts till 2022 were migrated to this platform (react+next+tailwind). While all efforts were made to migrate wihtout any loss, the migration lost some images and broke a bunch of links in old posts. If you spot anything amiss, please notify me?