Re: [Hutter Prize] Is the Hutter prize competition still taking place?

[James Bowery]

Yes. The competition is still underway.

On Fri, Feb 19, 2021 at 12:37 PM Austin Bryant <austin.b...@gmail.com> wrote:

I discovered the Hutter prize competition while researching and developing a lossless compression technique. 

...

Is the competition is still underway? 

[Sebastian Nerger]

Hello,

I'm not sure: is the time limit only for compressing.

Thank you

[Matt Mahoney]

100 hours to compress and 100 hours to decompress.

--
You received this message because you are subscribed to the Google Groups "Hutter Prize" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hutter-prize...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hutter-prize/8b9c8f73-6eb1-46f4-959f-19b8271c21a6n%40googlegroups.com.

[Sebastian Nerger]

Thank you Matt,

I hope my english is understandable. If I understand it correctly I only need to compress the file in 100 hours? Not 10 - OK. This seems not a problem, unlike than decompression.

But I am not(!) a math prof and in result of this I do the compression "in another way", but my PoC works.

So to optimize my code I have few questions:

- Can I use Java?

- Can you define / tell a minimum size, where I can stop the compression?

- How can I estimate my local performance against your test system? Is a file, standard compression (like gzip) and time on your system given for local run to calculate a ca.-factor for runtime?

Thank you

Bastie 

You received this message because you are subscribed to a topic in the Google Groups "Hutter Prize" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/hutter-prize/M-yFSexEklw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to hutter-prize...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hutter-prize/CACFhWPPPdF67%3DXFqJ2dPWxne2Avsej_u9LbYEQs_O9AG6UpD7w%40mail.gmail.com.

[Matt Mahoney]

The rules permit C++, Python, and x86 assembler. Personally I would allow Java. For testing, you have 100 hours on a machine with a geekbench score of 500. I suggest measuring your hardware with it and adjust accordingly. The compression ratio should not depend on your hardware or any input arguments.

To view this discussion on the web visit https://groups.google.com/d/msgid/hutter-prize/CAN85sfcVt4jfKgyZUaUGhrSi5M1VMG-q1RJA4JwnY75Sgn76EA%40mail.gmail.com.