Friday, September 5, 2014

Morse Learning Machine challenge!

A while ago I wrote about AG1LE's development of the ultimate CW decoder. Now Mauri took another approach. He wrote out a morse challenge, a competition to program a morse learning machine. If you like to learn about this challenge you better take a look on his blog.

AG1LE morse learning machine challenge
Kaggle Morse Learning Machine v1


Een tijdje geleden schreef ik over de ontwikkeling van een ultieme CW decoder door AG1LE. Nu benaderd Mauri het op een andere manier. Hij schreef een wedstrijd uit om een morse leer machine te programmeren, Als je meer wilt weten hier over kun je beter zijn blog lezen.


4 comments:

ag1le said...

Hi Bas

The Morse Learning Machine Challenge has really taken off much quicker than I anticipated. Within first 36 hours of the competition there has been 76 downloads of the dataset & example Python code.

We have two teams competing now, and within first 24 hours a very impressive *** one order of magnitude *** improvement in decoding accuracy vs. my benchmark was achieved. Aaron, KR1LLR is currently leading with score 1.65 (best possible score in this contest is 0.0).

I would like to get some European participation on this competition. Anybody interested in this challenge?

73
Mauri AG1LE

PE4BAS, Bas said...

Thanks for the extra information Mauri. I keep an eye on your blog with great interest. I wish I was clever enough to program. Luckely others are and I see myself only as someone that can test the software in realtime environment. It is good to read that first accomplishments have been made.

Hopefully someone here in Europe likes to participate. Good luck, 73, Bas

ag1le said...

This is a great opportunity for people who are interested in gaining knowledge about software development, data science, statistical data analysis, machine learning algorithms etc.

Kaggle community has been collaborating and solving some very hard problems - like Higgs Boson machine learning challenge - see the list of top challenges on http://www.kaggle.com/ front page.

In the first 2 1/2 days of this challenge there has been almost 100 downloads of the provided open source Morse decoder (Python based, there is also C++ code available for free), audio material and sample submission file.

I am doing my best to make it really easy to join this challenge and build a community and knowledge sharing among the participants. You have nothing to lose, but a lot knowledge to gain by joining this challenge.

The link to join is here: https://inclass.kaggle.com/c/morse-challenge


73
Mauri AG1LE




ag1le said...

We made it to ARRL news page this morning!

http://www.arrl.org/news/morse-learning-machine-challenge-catching-on-with-hams