3 ms·
Do you have some good reference about this hypothesis? I'm not really into neuroscience but this really interests me. I'm more on the machine learning side with
by albertzeyer 12y ago
Do you have some good reference about this hypothesis? I'm not really into neuroscience but this really interests me. I'm more on the machine learning side with artificial neural networks, but I often wondered whether some sleep-like phases could get me some improvements. Of course, the question is, what exactly should happen in that phase. What particular neural connections should I strengthen?
- coldtea 12y agoPerhaps it has no analogous to software neural networks? Looks like this "strengthening" is needed because there's actual "matter", "tissue" etc involved.
- mrexroad 12y agofrom what i recall (no documentation link, but it's been posted here before) strengthening is more about making associations/connections for the things that are important. so if you have CPU bound batch jobs that run each night for identifying new connections between clusters of data... it seems like it'd be pretty analogous. it's not like things we 'forgot' weren't stored, or somehow dropped from memory; we just couldn't access it when we needed to later, and once reminded we're fully aware. there's also been studies on people who suffer a traumatic experience; if they're deprived sleep following the event then the memories are less severe/traumatic.