Help
Options
Didn't know it?
click below
click below
Knew it?
click below
click below
Don't Know
Remaining cards (0)
Know
retry
shuffle
restart
0:00
Embed Code - If you would like this activity on your web page, copy the script below and paste it into your web page.
Normal Size Small Size show me how
Normal Size Small Size show me how
2236 week ten
| Term | Definition |
|---|---|
| concepts associative learning | contiguity (aristotle's principle) contingency (predictability) |
| contiguity (aristotle's principle) | two events are associated if they occur close together in time or space necessary for learning not sufficient |
| delay/trace conditioning | supports necessity |
| simultaneous/backward conditioning + blocking | contiguity alone doesn't guarantee learning |
| contigency (predictability) | degree to which the CS predicts the US |
| rescorla: p(us/cs) > p(us/no cs) | -> learning happens |
| levels of contingency | positive, zero and negative |
| positive | CS predicts US -> excitatory conditioning |
| zero | CS does not predict US -> no learning |
| negative | CS predicts absence of US -> inhibitory conditioning |
| Rescorla classic experiment | test whether contingency not just contiguity determines learning |
| Tones (CS) and shocks (US) | all rats had same cs-us pairing rate (0.4) shock probability without tone varied |
| stronger learning when was unique predictor of shock | learning depends on predictive value of CS, not just frequency |
| Rescorla Wagner Model | change in associative strength salience of cs salience of us actual outcome (maximum associative strength) expected outcome (sum of current associative strengths) |
| learning | driven by prediction error |
| big surprise | more learning |
| no surprise | no learning |
| asymptote | learning levels off as CS fully predicts US |
| RW model | acquisition, blocking, extinction, conditioned inhibition and overexpectation |
| acquisition | learning occurs fastest early, then slows |
| blocking | prior learning blocks new CS from gaining strength |
| extinction | when CS occurs without CS, association weakens |
| conditioned inhibition | CS predicts absence of US |
| overexpectation | combined CSs over predict US -> both weakened |
| neural | dopamine and prediction error |
| dopamine neurons respond | to reward prediction errors |
| unexpected reward | spike in firing |
| expected reward | no spike |
| expected but absent reward | dip in firing |
| attentional theories | focus on attention to CS if CS is uninformative, attention to it drops -> slow learning explains latent inhibition (pre exposure effect) which R-W cannot |
| comparator theories | focus on performance not learning per trial emphasise the context and comparison between CS and background cues |
| blocking | performance issue, not a learning failiure |
| can explain unblocking | learning masked by stronger CS |
| contiguity | necessary, not sufficient |
| contingency | critical - learning depends on how well CS predicts US |
| prediction error | learning happens when there's a mismatch between expected and actual outcomes |
| rescorla wagner | powerful model for quantifying learning, but has limitations |
| alternative theories | emphasise attention (CS effectiveness) and performance (Comparator) |
Created by:
brendonpizarro1