5 ms·

How would this apply to learning a conceptual topic, as opposed to a physical skill?

by halpme 11y ago

bbhill 11y ago
I think problem sets in textbooks already do this; you do multiple problems relating in different ways to a lesson and so you learn how to master it in different ways.
- halpme 11y ago
  That sounds a lot like deliberate practice, which is an idea that's been around for a while.
guimarin 11y ago
when trying to learn something, engage with it repeatedly on exponential decay. IE. Index cards. Go through them as many times as it take to learn whatever you are learning. Then perform a 20-30 minute context switch. do again. Now wait 1 hour. Do the cards again. Wait 2 hours. Again. Do this until you review those cards at the week level. You can learn almost anything with very high recall 6-9 months later with this method. edit: Sleep is an extremely important component of this. You must be well rested, and preferably do not ingest significant amounts of alcohol before sleeping as that will mess up short term to long term memory mapping from hippo to pfc.
- eicossa 11y ago
  is there any application or software that helps to do this ? for instance, reading dense technical documentation (like opencv or scikit-learn) or papers ? . i currently try to approximate this process using a hap-hazard combination of emacs, org-mode and the org-drill packages. are there other better ways out there ? . Don't say anki, because anki seems to only be good for memorizing and engaging with material that seems to have a well-defined schema (eg. vocabulary, rules of grammar, etc) but not for applying the scheme you mentioned to documents of arbitrary complexity.
- spangry 11y ago
  I can anecdotally attest to this. When I had a piano recital coming up I used to practice very intensely, but would 'hit a wall' at some point. When I came back to it the next day I'd find my playing had improved significantly. It almost felt like the gains from practising got stored up somewhere while I was awake, and slowly unspooled while I was sleeping.
  - guimarin 11y ago
    this is how short-term -> long-term memory works. You build up 'competence; in the hippocampus during waking hours, and then when you sleep 'a part' of those memories are shifted into other areas of the brain so the hippo can be fresh for the next day. Caveats apply to things sometime take days, don't move, get erased etc. hippo is very capable so people can operate without sleep for quite a while but eventually overflowing this buffer/registry causes insanity/death. hilariously, your long-term appreciation of time is just the working set in hippo divided by the stored area in long term memory. thus as you get older time appears to go by faster in the aggregate. experience of simple in the moment phenomena is controlled elsewhere.

How would this apply to learning a conceptual topic, as opposed to a physical skill?

by halpme 11y ago

How would this apply to learning a conceptual topic, as opposed to a physical skill?

I think problem sets in textbooks already do this; you do multiple problems relating in different ways to a lesson and so you learn how to master it in different ways.

halpme 11y ago
That sounds a lot like deliberate practice, which is an idea that's been around for a while.

guimarin 11y ago

when trying to learn something, engage with it repeatedly on exponential decay. IE. Index cards. Go through them as many times as it take to learn whatever you are learning. Then perform a 20-30 minute context switch. do again. Now wait 1 hour. Do the cards again. Wait 2 hours. Again. Do this until you review those cards at the week level. You can learn almost anything with very high recall 6-9 months later with this method. edit: Sleep is an extremely important component of this. You must be well rested, and preferably do not ingest significant amounts of alcohol before sleeping as that will mess up short term to long term memory mapping from hippo to pfc.

eicossa 11y ago
is there any application or software that helps to do this ? for instance, reading dense technical documentation (like opencv or scikit-learn) or papers ? . i currently try to approximate this process using a hap-hazard combination of emacs, org-mode and the org-drill packages. are there other better ways out there ? . Don't say anki, because anki seems to only be good for memorizing and engaging with material that seems to have a well-defined schema (eg. vocabulary, rules of grammar, etc) but not for applying the scheme you mentioned to documents of arbitrary complexity.
spangry 11y ago
I can anecdotally attest to this. When I had a piano recital coming up I used to practice very intensely, but would 'hit a wall' at some point. When I came back to it the next day I'd find my playing had improved significantly. It almost felt like the gains from practising got stored up somewhere while I was awake, and slowly unspooled while I was sleeping.
- guimarin 11y ago
  this is how short-term -> long-term memory works. You build up 'competence; in the hippocampus during waking hours, and then when you sleep 'a part' of those memories are shifted into other areas of the brain so the hippo can be fresh for the next day. Caveats apply to things sometime take days, don't move, get erased etc. hippo is very capable so people can operate without sleep for quite a while but eventually overflowing this buffer/registry causes insanity/death. hilariously, your long-term appreciation of time is just the working set in hippo divided by the stored area in long term memory. thus as you get older time appears to go by faster in the aggregate. experience of simple in the moment phenomena is controlled elsewhere.