Thu 19th July - Learning Flashcards
Refer back to this when doing lab 3 & 4
Learning outcomes:
Demonstrate an understanding of habituation and sensitisation.
Gain knowledge of classical conditioning, show knowledge of the terminology, and be able to provide examples of this in practice.
Demonstrate an understanding of the principles and application of operant conditioning, including terminology. Be able to describe concepts, using examples.
Have knowledge of how taste aversion can develop and its adaptive value.
Understanding of how learning capabilities, e.g., spatial learning, may be adaptive to particular environments.
http://www.youtube.com/watch?v=9T1vfsHYiKY
Example: Sirocco and misdirected sexual behaviour – training.
http://www.youtube.com/watch?v=89GQPfWas78
Positive reinforcement - gets a treat everytime he interacts with the ‘bird’ like dummy
Why is Learning Important?
Learning as defined by Thorpe (1963):
“that process which manifests itself by adaptive changes in individual behaviour as a result of experience”. (Manning & Dawkins, 2012, p.240.)
Animals live in dynamic environments
• Adaptive – individuals retain responses useful to them.
– Offspring won’t inherit learned responses (general ability to perform them), but social learning can come into play.
• Equips animals with ability to adjust their behaviour, benefit from experience.
– Behavioural flexibility.
– Advantages: avoiding danger, more efficiently finding resources.
Simple Learning
Two simple learning processes help animals interact within environments more efficiently.
• Habituation.
• Sensitisation.
Simple Learning
Habituation = harmless stimulus originally elicits a response, but response gradually decreases over time.
This learning is stimulus-specific.
‒ E.g., fireworks, car backfiring, door slamming.
E.g., three-spined sticklebacks (Peeke & Veno, 1973):
‒ Males est. territories, tested aggressive responses to familiar and unfamiliar males.
‒ The more novelty = stronger response.
Habituation
• Individual variation in speed of this learning.
– E.g., goslings habituate slowly to threatening predators.
Important to establish it is learning – not sensory adaptation or
fatigue.
The initial responses can recover (if no exposure to stimulus for some time).
Question: what could be some costs to animals if they did not habituation to various stimuli in their environment?
-Stress
-Wasted energy
-Pray species reacting to every russle in the grass would miss out on social interactions and grazing, bad daily life quality.
-Hyper vigalence
-Reduced mating
-Predators chasing everything
• Hypervigilance to potential predators. – Reduced feeding time.
Predators chasing everything.
Reduced parental care.
Energy expenditure.
Trade off – reacting (and costs) vs. predation risk.
Safe habituation – discrimination between dangerous and harmless events.
Challenging in noisy environments.
– Other strategies.
Simple Learning
Sensitisation = increased responding to a stimulus.
• Affected by intensity – greater intensity = greater sensitisation.
First studied ‘accidentally’ by Pavlov.
Classical conditioning AKA Pavlovian/respondent conditioning.
• Noticed salivation in dogs before they touched the meat powder.
Classical conditioning =
bringing the elicitation of involuntary responses under control of a stimulus that previously had no association with that response.
Classical Conditioning
Involuntary behaviour:
Food on tongue … salivation.
Puff of air in eye … blinking.
Bright light/darkness … changes in pupil constriction/dilation.
Loud noise/sudden motion … startle response.
Classical Conditioning
Example: unconditioned response – startle.
http://www.youtube.com/watch?v=O4rfQSgkZOE
Uncontrolled reflex response from mother to stimulus
Unconditioned stimulus (US) = has intrinsic significance to an animal.
Unconditioned response (UR) = involuntary response (AKA reflex) the US elicits.
US-UR relationships is not learned.
Neutral stimulus (NS) = no significance.
US presented + NS, repeatedly …
The previously-NS begins to elicit the same UR.
Conditioned response (CR) = when the UR elicited by just CS.
Classical Conditioning
Temporal contiguity is important for this conditioning to occur.
Arrangement of stimuli in time.
E.g., CS presented first, then the US – with some overlap.
Operant Conditioning
Operant conditioning = when the reoccurrence of voluntary behaviours is increased or decreased, depending on the consequences that followed that behaviour.
Voluntary behaviours:
Recall.
Sit, down.
Stationing.
Opening mouth.
Lifting leg.
Manipulating items.
Reinforcement = increases.
Punishment = decreases.
Positive = addition.
Negative = removal.
Reinforcement = increases.
Punishment = decreases.
Positive = addition.
Negative = removal.
Questions
What are these examples of?
What behaviour will likely reoccur?
You call a dog, they come to you, you give them a treat.
Positive reinforcement to dog
Adding treat = Stimulus
Behaviour = Dog returning
Behaviour more likely to occur
A rat approaches a trap, the trap goes off (doesn’t trap the rat)
Behaviour = Approaching trap
Stimulus = trap going off
Positive punishment
Rat will avoid trap
An animal is being handled and it wriggles free.
Behaviour = wriggaling free
Animal free = handling aka stimulus removed
Negative reinforcement because something is removed
Animal will continue to wriggle next time due to the success of escaping the handler
A cub claws at their mother and she gets up and walks away.
Negative punishment is mother walking away from cub
Application: use of unconditioned and conditioned positive reinforcers.
Unconditioned reinforcers = no learning required to be reinforcing.
• E.g., food, warmth, etc.
Conditioned reinforcers = attain reinforcing value via classical
conditioning.
• E.g., verbal praise, clickers.
Temporal contiguity is also important here.
CLICK AND THEN TREAT
Operant Conditioning
Shaping = reinforcing successive approximations of the final response.
• Animal initially received reinforcement for any small part of the
final desired behaviour.
– E.g., a slight turn of the head.
• You gradually shift the criteria for when you deliver reinforcement.
– E.g., master a slight head turn (9/10 trials), wait for a 45° turn and reinforce only for that, then a 90° turn, then head turn + shoulder turn, etc.
• Often used without realising.
– E.g., training an animal to jump higher.
You will practice shaping in Lab 3.
http://www.youtube.com/watch?v=1iKaNN_XrnE
Shaping training to get the dog to bow
Birds in the UK learnt to peck through foil milk lids to get milk.
Behaviour learnt through positive reinforcement
(milk is the reward)
Taste Aversion
Taste aversion = when an animal becomes unwell after eating a food and then avoids that food.
Learning, classical conditioning basis, but a bit different:
Learning can take place with long delays (~12 hours).
Learning can occur in just one trial.
Taste Aversion
Constraints on learning seen in taste aversion.
Example: taste aversion in quails vs. rats.
Both ingested blue salty water, then made ill.
Then given choice:
– Non-coloured salty water OR
– Blue non-salty water.
Quails: avoided blue water.
Rats: avoided salty water.
Question: why might this be?
Think about how these animals select food:
Quails: blue water = visual cue.
Rats: salty water = taste and smell.
Different sensory cues for selecting/avoiding food
Taste Aversion
Other points:
Nausea appears important for conditioning taste aversion. – Cf.electricshock,orparalysisofspinalneurons.
Novel tastes/settings are more easily conditioned than familiar ones.
What adaptive value is there in this type of learning?
• Survival.
Spatial Learning
Example: seed caching in birds – long-term, spatial memory.
These birds have differing anatomies.
Clark’s nutcrackers hide up to 33,000 seeds.
– Pouch for transporting seeds to cache sites.
Pinyon jays cache up to 20,000 seeds.
– Expandable oesophagus for carrying seeds.
Scrub jays & Mexican jays cache far less.
– No specialised anatomy.
Spatial Learning
Example: corvids tested on two learning/retention tasks:
1 (A) = spatial learning task (remember location).
– Clark’s nutcrackers were significantly more successful.
2 (B) = non-spatial learning task (remember colours).
– Pinyon and Mexican jays were top performers.
