Top all-time · #10763
content
Personal blog · charlesmartin14.wordpress.comaidata
content is a personal blog at charlesmartin14.wordpress.com about ai and data. Since 2013, 7 different Hacker News readers have linked it 10 times, most often “Why Deep Learning Works II: the Renormalization Group”. It is still writing: the last post was on 2025-12-13, and you can follow it by feed.
Too few sampled comments (6) to say how often HN recommends it.?
Why HN links it
HN readers link these posts to explain why deep learning works, using spin models and renormalization group ideas, and to describe the Bitcoin crash as self-organized criticality.
Summarised by Claude from a sample of 6 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Thoughts on Data Science, Machine Learning, and AI
Most-linked posts?
- Why Deep Learning Works II: the Renormalization Group2 people · 2 mentions · first linked 2015-07-05
- Why does Deep Learning work?2 people · 2 mentions · first linked 2015-03-23
What Hacker News says?
Deep Neural Networks learn layers of linear functions - represented by adjustable 'learning' weighted connections. Each layer of neuron activations is a new representation of the data - produced by the weighted connections. Function compositions of linear functions are still only a linear function. Each neuron in a layer sums it's weighted inputs this summation is a non-linearity that allows layers to be composed -…
Comment linking /2015/04/01/why-deep-learning-works-ii-the-renormalization-group/
on “Machine learning works spectacularly well, but mathematicians aren’t sure why”Submitted Why Deep Learning Works II: the Renormalization Group as a story · 120 points
HN discussionThis is as good a post as any about the theory of why deep learning works so well - esp. if you know a little about spin models from statistical mechanics https://charlesmartin14.wordpress.com/2015/02/26/why-does-de... additional flavor: http://arxiv.org/pdf/1412.0233.pdf and slides 16-22 of https://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsd...
Comment linking /2015/02/26/why-does-deep-learning-work/
on “Deep Learning vs. Machine Learning vs. Pattern Recognition”Submitted Why does Deep Learning work? as a story · 3 points
HN discussionIt is difficult to understand the implications of these assumptions and if they really apply to supervised deep learning nets. It has been know for a very long time that simple models, like the Random Energy Model, displays a spin glass transition at low temperature. (the REM is a p-infinite limit the mean field p-spherical spin glass used in the earlier papers by LeCun) So it is expected that a random network may a…
Comment linking /2015/03/25/why-does-deep-learning-work/
on “Deep Learning Without Poor Local Minima”Submitted Cloud-crawler, a distributed Ruby DSL for crawling the web via Amazon ec2 as a story · 2 points
HN discussion'Technically' this is correct--the RBM CD algo is not minimizing this function; that's not the point. It is known that when training an RBM, the reconstruction error decreases but not monotonically; in fact it fluctuates. In the words of Hinton, 'trust it but don't use it'. http://www.cs.toronto.edu/~hinton/absps/guideTR.pdf (which is cited in the post as well) So in a global sense, yes, I would say that the RBM doe…
Comment linking /2015/01/16/the-bitcoin-crash-and-how-nature-works/
on “Why Deep Learning Works II: the Renormalization Group”The Bitcoin crash is similar to crashes in other financial markets and is a signature of self organized criticality: https://charlesmartin14.wordpress.com/2015/01/16/the-bitcoin...
Comment linking /2015/01/16/the-bitcoin-crash-and-how-nature-works/
on “Amid Bitcoin's Bloodbath, Silence from Silicon Valley Press”See my blog: https://charlesmartin14.wordpress.com/2015/04/01/why-deep-le... I try to highlight the paper and some of the history and relevance, in a general way, but also hitting the math hard. the more general idea, which I am still formulating, is that DL systems are very different from traditional ML (ala VC theory) In traditional ML systems, we tune the regularizer to optimize the capacity of the learner In dee…
Comment linking /2015/03/25/why-does-deep-learning-work/
on “Exact mapping between Variational Renormalization Group and Deep Learning (2014)”