definition of learning: learning refers to the relatively permanent change in a subject’s behavior...
TRANSCRIPT
Definition of Learning:
Learning refers to the relatively permanent change in a subject’s
behavior to a given situation brought about by repeated experiences in that situation, provided that the behavior
change cannot be explained by instinct, maturation, or temporary
states.
Are the examples given, examples of learning?1. The cessation of thumb sucking by an infant.
– Due to maturation
2. The acquisition of language in children.– Learning
3. A computer program generates random opening moves for its first 100 chess games and tabulates the outcomes of those games. Starting with the 101st game, the computer uses those tabulations to influence its choice of opening moves.– Learning
4. A worm in places in a T maze. The left arm of the maze is brightly lit and dry; the right arm is dim and moist. On the first 10 trials, the worm turns right 7 times. On the next 10 trials, the worms turns right all 10 times.
- Learning
Are the examples given, examples of learning?5. Ethel stays up late the night before the October GRE administration and consumes
large quantities of licit and illicit pharmacological agents. Her combined score is 410. The night before the December GRE administration, she goes to bed early after a wholesome dinner and a glass of milk. Her score increases to 1210. Is the change in scores due to learning. Is the change in pretest regimen due to learning?– Change in score: Due to temporary state. Change in pretest regimen: Learning
6. A previously psychotic patient is given Dr. K’s patented phrenological surgery and no longer exhibits any psychotic behaviors.- Change in behavior due to surgery, not repeated experiences
7. A lanky zinnia plant is pinched back and begins to grow denser foliage and flowers.
- Due to maturation
Are the examples given, examples of learning?
8. MYCIN is a computer program that does a rather good job of diagnosing human infections by consulting a large database of rules it has been given. If we add another rule to the database, has MYCIN learned something? - Debatable
9. After pondering over a difficult puzzle for hours, Jane finally figures it out. From that point on, she can solve all similar puzzles in the time it takes her to read them.- Learning (if pondering = trial & error and other problems solving techniques)
10.After 30 years of smoking two packs a day, Zeb throws away his cigarettes and never smokes again.- Learning: repeated experiences (smoking) and their negative effects (breathing, cancer, bad breath)
Associations
Pavlov
Unconditioned Stimulus (UCS)
Unconditioned Response (UCR)
Conditioned Response (CR)
Conditioned Stimulus (CS)
Acquisition Extinction
Generalization
DiscriminationSpontaneous Recovery
Classical Conditioning: the process of learning the association between two stimuli.
Classical Conditioning: Pavlov’s Dogs
UCS UCR
UCS
CS
CS CR
Boy, do Iever salivate
a lot!
Boy, do Iever salivate
a lot!
What’s wrong with me? Now I’m salivating at
a bell!
Before
During
After
UCR
An MIT student spent an entire summer going to the Harvard football field every day wearing a black and white striped shirt, walking up and down the field for ten or fifteen minutes throwing birdseed all over the field, blowing a whistle and then walking off the field. At the end of the summer, it came time for the first Harvard home football game, the referee walked onto the field and blew the whistle, and the game had to be delayed for a half hour to wait for the birds to get off of the field. The guy wrote his thesis on this, and graduated.
Unconditioned Stimulus? Birdseed
Unconditioned Response?
Fly to field & eat
Conditioned Stimulus?
Black & White Striped Shirt
Conditioned Response?
Fly to field to eat
The initial learning process. The pairing of the CS with the UCS. Learning takes place when the CS is presented 1/2 second prior to the UCS.
The tendency, once a response has been conditioned, for stimuli similar to the conditioned stimulus to produce a similar response.
CR produced in response to specific stimuli and not to other stimuli that do not signal an UCS.The ability to distinguish between similar CS.
Bald Gentlemen vs. Skin heads; Pit Bull vs. Poodle
A decrease in the CONDITIONED RESPONSE that occurs when the UNCONDITIONED STIMULUS no longer follows the CONDITIONED STIMULUS.
SPONTANEOUS RECOVERY: the reappearance of the extinct conditioned response in the presence of the conditioned stimulus
Applications of Classical ConditioningIs human behavior nothing more
than a bunch of conditionedbehaviors?
John Watson
Case of “Little Albert”
UCSUCSLoud Noise
CSCSWhite Rat
UCRUCRFear
CRCRFear
Applications of Classical Conditioning
Case of “Little Albert”
Five days later Albert showsgeneralizationgeneralization to other white,
furry objects.e.g. Baby Albert was conditioned to fear a white rat,
but also feared cotton balls, rabbits, white sweaters, etc.
Extinction:After a period of time passes when CS is not paired with
UCS, CS returns to being an NS
e.g. Baby Albert would eventually cease to be afraid of white
fluffy things after they were not paired with
a horrible and frightening noise
Do cognitive processes and biological constraints affect
classical conditioning?
Rescorla and Wagner (1972): animals learn to “expect” an unconditioned stimulus; this shows cognition at work: the animal learns the predictability of a second associated event after the firstConditioning an alcoholic with a nauseating drink might not work because they are “aware” of what causes the nausea---the drink, not alcohol.
Martin Seligman found that dogs given repeated shocks with no opportunity to avoid them developed a passive resignation called learned helplessness. In new situation, animals that escaped the first shocks learned personal control and were able to able to easily escape shocks thereafter.
Cognition
Biology vs. Environment Garcia and Koelling: -Animals can learn to avoid a drink that will make them sick, but not when it’s announced by a noise; so Pavlov was wrong in claiming that any stimulus could serve as a conditioned stimulus.-We are biologically prepared to learn certain associations and not others: we learn to fear snakes, but not flowers-Taste aversions result from biology. The smell and taste of a nauseating food become the CS for sickness.-Secondary disgust: an aversion that reminds a person of something that is considered disgusting in its own right. Ex. You won’t eat a piece of chocolate formed like dog feces.
Ivan Pavlov-His Legacy
• Classical Conditioning principles can be applied to all organisms, from earthworms to people
• Measurable, objective procedures…not a guessing game
Applications of CC
• Health (ex: drug use, alcoholism)
• John Watson’s Little Albert Experiment: Led to the question-Can fears be extinguished?
• Rape-People & Locations serving as Conditioned Stimuli leads to a Conditioned Fear response
• Advertising
Types of Conditioning
ClassicalClassicalProcess of associating
two stimuli
ClassicalClassicalProcess of associating
two stimuli
OperantOperantProcess of associating
a response & itsconsequence
OperantOperantProcess of associating
a response & itsconsequence
Lightening
Thunder
Pulling candy machinelever
Delivery of candy bar
During which type of conditioning does an organism learn associations between events that it doesn’t control?
During which type of conditioning does an organism learn associations between it’s own behavior and resulting events?
CLASSICAL CONDITIONING
OPERANT CONDITIONING
Thorndike: Law of Effect-rewarded behavior is likely to reoccur.
Skinner: Skinner Box & “behavioral technology”
Shaping – reinforcers guide behavior toward closer approximations of a desired goal.
Successive Approximations-method used to shape organisms
ShapingDemonstration
Principles of Reinforcement
ReinforcerReinforcer - any event that increases the frequencyof the preceding event
Positive ReinforcersPositive ReinforcersIntroduce (+) stimulus
(e.g., food)
Negative ReinforcersNegative ReinforcersRemove (-) stimulus(e.g., electric shock)
Reinforcers ALWAYS strengthen behavior!Reinforcers ALWAYS strengthen behavior!
More Reinforcement
Primary ReinforcersPrimary ReinforcersInnately satisfying,
Not learned(e.g., getting food)
Secondary ReinforcersSecondary ReinforcersAssociated with primary
reinforcers & learned(e.g., praise)
Reinforcers ALWAYS strengthen behavior!Reinforcers ALWAYS strengthen behavior!
Still More Reinforcement
Immediate ReinforcementImmediate ReinforcementReinforce immediately preceding
behavior (e.g., nicotine)
Delayed ReinforcementDelayed ReinforcementReinforcement at some point
after behavior occurs(e.g., paychecks)
Reinforcers ALWAYS strengthen behavior!Reinforcers ALWAYS strengthen behavior!
STIMULUS RESPONSE REINFORCEMENT/PUNISHMENT
Something thatHappens TO an
organism
Behavior performedBy the person or
Animal.
Types of Reinforcers:
1. Positive-something good happens & strengthens behavior
2. Negative-something bad is taken away & strengthens behavior
3. Primary-pleasant in their own right-food (positive), being relieved of pain (negative)
4. Secondary (also called Conditioned)-is associated with a primary reinforcer (classical conditioning)
5. Immediate –reinforcer that immediately follows the response-ex: drugs
6. Delayed -reinforcer that is postponed to well after the desired behavior-ex: long-term reward (safe sex vs. immediate sex)
Reinforcers ALWAYS strengthen behavior!Reinforcers ALWAYS strengthen behavior!
PUNISHMENT-decreases frequency of preceding behavior
•Effective When? Strong, Immediate, Consistent
•Undesirable When? Increases in aggression, fear (avoidance behavior), low self esteem, depression, performance of behavior in “safe” conditions, no guidance toward desirable behavior.
VARIABLE RATIO - reinforcement after a random # of responses. Greater resistance to extinction, but initial learning is slower than continuous reinforcement. Ex: Gambling.
FIXED RATIO-reinforcement after specific # of responses. Ex: Piece-work
FIXED INTERVAL -reinforcement after a specific amount of time. Ex: Checking the mail.
VARIABLE INTERVAL-reinforcement after random amount of time. Most resistant to extinction. Ex: Pop Quizzes
REINFORCEMENT
SCHEDULES
Latent Learning-learning that is not apparent until a reward is presented to demonstrate the learning.
Rat that is never reinforced continues to make many errors in completing the maze.
Rat that explored the maze for 10 days, then given a reward on Day 11, shows signs of Latent Learning.
Rat that is always reinforced gives evidence of learning throughout trials.
Tang & Hall Study (1995) Hey kids! I’ll pay you to
play with that toy
This toy is great-I’m
going to play with it all the
time!
“If I’m not getting paid, why bother.” OR “If I have to be bribed, then it must not be that great of a toy.”
OVERJUSTIFICATION EFFECT-decreasing intrinsic motivation for a behavior after an extrinsic reward is provided.
Salary
Performance
Biology & Operant Conditioning
• Naturally adaptive behaviors are easier to condition (example: Pigeons avoid shock by flapping wings but slower to learn to flap wings for food.)
• “Instinctive Drift”- animals reverting back to predisposed behavior rather than conditioned behavior.
Skinner’s Legacy
• Give rewards that promote desired behavior
• Personal freedom? Not in Skinner’s world.
Operant Conditioning Applications
School Work HomeComputers &
Immediate Reinforcement
Increase Productivity-Profit Sharing
Parenting-reward good behavior, ignore whining
On-line Testing/Interactive
Software
Rewards work when behavior is well-
defined and obtainable
Goal setting-reward your own desired behaviors
Sports & Incremental Reinforcement
Praise can be an effective managerial
technique
Spending Behavior
Albert Bandura’s Bobo-doll experiment: Aggressive behaviors learned through observation of a model.