[LCWO LOGO]  

Login

User name:
Password:


Language
Български Português brasileiro
Bosanski Català
繁體中文 Česky
Dansk Deutsch
English Español
Suomi Français
Ελληνικά Hrvatski
Magyar Italiano
日本語 한국어
Bahasa Melayu Nederlands
Norsk Polski
Português Română
Русский සිංහල
Slovenščina Srpski
Svenska ภาษาไทย
Türkçe Українська
简体中文
Who is online? (30)


LCWO Discussion Forum [Atom LCWO Forum Feed]

This is a simple discussion forum for LCWO users. Feel free to use it for any kind of discussion related to this website.

Thread: Morse Machine

Back to the Forum

AuthorText


Posted: 2014-06-30 02:28
I'd like to suggest a change on Morse machine to better train on missed characters. At present the count on missed characters is only increased by a small amount. As a result, the missed character is often not encountered for a while. Could we have a CW setting that we can set to the amount to increase a missed or requested repeat character?

Also, is it possible to measure the time it takes for us to respond and use that to adjust the character frequency? Therefore characters that we take a long time to respond would increaase in frequencyand help s learn them.

Thanks for considering hese ideas
Administrator


Posted: 2014-06-30 08:24
Hi John,

good ideas.. another idea to make learning more efficient would be using a spaced repetition algorithm (http://en.wikipedia.org/wiki/Spaced_repetition) like one of the SuperMemo algorithms, but I have not looked into this closely yet.

In any case, I don't know if you already found out (and it is not a documented feature yet :), if you want to do quick manual adjustments to the error bars in the MorseMachine, you can just click on them to change the values.

73
Fabian


Posted: 2014-07-01 02:20
I just found the manual adjustment feature for the error bars. Very useful and almost makes my suggestions not needed.
Thanks again for making THE BEST tool for learning CW (or re-learning in my case).
I passed the 5 wpm several years ago but never made any CW QSO's. You have rekindled my interest in CW!


Posted: 2014-07-01 02:39
I just looked at the Spaced_repetition algorithm. It looks very similar to adjusting the error bars but with less granularity.

One other thing that comes to mind is the training methods for Neural Networks to avoid "over-training":
http://en.wikipedia.org/wiki/Overfitting

where a Neural Network can only recognize the "perfect" examples presented during training.

This means that we should emphasize variations in the training set i.e. variation in tone (you already do this), noise in the signal, signal fade, slight changes in mark/space timing, etc. to better prepare for real world CW.

But we still don't know how close human learning of CW parallels the training of a Neural Network?
Probably pretty close. Fascinating stuff.

Back to the Forum

You must be logged in to post a message.