[tdr2d]

Obviously, the code may be elegant and compact, 78% accuracy is considered very very bad for MNIST.

A dummy model written with Tensorflow easilly reaches 90% accuracy. The best models ranked at 99,87%, see the benchmark : https://paperswithcode.com/sota/image-classification-on-mnis...

[esafak]

The article emphasizes the wrong thing, in my view. The interesting part is that compression -- without learning a model -- can be used for classification. This raises the question of what other information-theoretic measures can be used; cheaper, lossy ones.

To Compress or Not to Compress- Self-Supervised Learning and Information Theory: A Review https://arxiv.org/abs/2304.09355\*

[anitil]

I remember seeing an example of using zip to classify languages. You take a set of documents of equal size where you know the languages, then individually concatenate and zip them with the unknown text. The smallest compressed output is likely to be the target language.

I can't find the original blog, but there's a note about it here - https://stackoverflow.com/questions/39142778/how-to-determin...

[Quenty]

I mean, they’re using knn underneath. You can apparently get 97% accuracy with normal knn, at n=4 if you compare pixel distance.

So another way to frame this might be that gzip costs a lot of accuracy but may lead to better performance.

https://newpblog.netlify.app/2018-01-24-knn-analysis-on-mnis...

[a_wild_dandan]

An increasingly common refrain in machine learning is “intelligence is compression.” Folks who believe that might bristle at the distinction between learning and compression.

[sundarurfriend]

The point is not to have "elegant and compact" code, this is meant to be a fun curiosity, and doing it in 10 lines is just an additional layer of challenge for the heck of it.

The interesting thing is not in whether GZip can achieve SOTA, it's that it can do a decent job at all. (The interesting thing is not in whether the bear can recreate Mozart exactly, it's that it can play the piano at all.)

[ActivePattern]

Yeah, it does demonstrate that you can use compression to measure similarity of two images.

But it also demonstrates that it's a pretty poor similarity measure. Something as simple as counting % of matches between the black and white pixels performs much better.

[m00x]

It's not trying to break records, it just shows a neat aspect of compression. It's still 8 times better than baseline, which showcases that compression can learn representation.