Definition of Learning: - PowerPoint PPT Presentation

1 / 37
About This Presentation
Title:

Definition of Learning:

Description:

Title: PowerPoint Presentation Last modified by: Olexiy Created Date: 10/14/1999 3:19:44 PM Document presentation format: On-screen Show (4:3) Other titles – PowerPoint PPT presentation

Number of Views:135
Avg rating:3.0/5.0
Slides: 38
Provided by: wikis2128
Category:

less

Transcript and Presenter's Notes

Title: Definition of Learning:


1
Definition of Learning
Learning refers to the relatively permanent
change in a subjects 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.
2
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
  • 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

3
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.
    Ks 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

4
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)
  • 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)

5
Classical Conditioning
Acquisition
Extinction
Discrimination
Spontaneous Recovery
Associations
Conditioned Stimulus (CS)
Unconditioned Stimulus (UCS)
Generalization
Conditioned Response (CR)
Pavlov
Unconditioned Response (UCR)
6
Classical Conditioning the process of learning
the association between two stimuli.
7
Classical Conditioning Pavlovs Dogs
Boy, do I ever salivate a lot!
Before
UCS
UCR
CS
During
Boy, do I ever salivate a lot!
UCR
UCS
Whats wrong with me? Now Im salivating at a
bell!
After
CR
CS
8
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.
9
M.I.T. Example
Unconditioned Stimulus?
Birdseed
Unconditioned Response?
Fly to field eat
Conditioned Stimulus?
Black White Striped Shirt
Conditioned Response?
Fly to field to eat
10
ACQUISITION
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.
11
The tendency, once a response has been
conditioned, for stimuli similar to the
conditioned stimulus to produce a similar
response.
GENERALIZATION
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
DISCRIMINATION
12
EXTINCTION
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
13
Applications of Classical Conditioning
Is human behavior nothing more than a bunch of
conditioned behaviors?
Case of Little Albert
UCS Loud Noise CS White Rat
UCR Fear CR Fear
John Watson
14
Applications of Classical Conditioning
Case of Little Albert
Five days later Albert shows generalization 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
15
Do cognitive processes and biological constraints
affect classical conditioning?
16
 
Cognition
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 first Conditioning 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.
17
Biology vs. Environment
Garcia and Koelling -Animals can learn to avoid
a drink that will make them sick, but not when
its 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 wont eat a piece of chocolate formed like
dog feces.
18
Ivan Pavlov-His Legacy
  • Classical Conditioning principles can be applied
    to all organisms, from earthworms to people
  • Measurable, objective proceduresnot a guessing
    game

19
Applications of CC
  • Health (ex drug use, alcoholism)
  • John Watsons 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

20
Types of Conditioning
Classical Process of associating two stimuli
Operant Process of associating a response
its consequence
Lightening
Pulling candy machine lever
Thunder
Delivery of candy bar
21
During which type of conditioning does an
organism learn associations between events that
it doesnt control?
CLASSICAL CONDITIONING
During which type of conditioning does an
organism learn associations between its own
behavior and resulting events?
OPERANT CONDITIONING
22
Thorndike Law of Effect-rewarded behavior is
likely to reoccur.
Skinner Skinner Box behavioral technology
23
Shaping reinforcers guide behavior toward
closer approximations of a desired goal.
Successive Approximations-method used to shape
organisms
Shaping Demonstration
24
Principles of Reinforcement
Reinforcer - any event that increases the
frequency of the preceding event
Positive Reinforcers Introduce ()
stimulus (e.g., food)
Negative Reinforcers Remove (-) stimulus (e.g.,
electric shock)
Reinforcers ALWAYS strengthen behavior!
25
More Reinforcement
Primary Reinforcers Innately satisfying, Not
learned (e.g., getting food)
Secondary Reinforcers Associated with
primary reinforcers learned (e.g., praise)
Reinforcers ALWAYS strengthen behavior!
26
Still More Reinforcement
Immediate Reinforcement Reinforce immediately
preceding behavior (e.g., nicotine)
Delayed Reinforcement Reinforcement at some
point after behavior occurs (e.g., paychecks)
Reinforcers ALWAYS strengthen behavior!
27
STIMULUS ? RESPONSE ? REINFORCEMENT/PUNISHMENT
Reinforcers ALWAYS strengthen behavior!
Something that Happens TO an organism
Behavior performed By the person or Animal.
  • Types of Reinforcers
  • Positive-something good happens strengthens
    behavior
  • Negative-something bad is taken away
    strengthens behavior
  • Primary-pleasant in their own right-food
    (positive), being relieved of pain (negative)
  • Secondary (also called Conditioned)-is associated
    with a primary reinforcer (classical
    conditioning)
  • Immediate reinforcer that immediately follows
    the response-ex drugs
  • Delayed -reinforcer that is postponed to well
    after the desired behavior-ex long-term reward
    (safe sex vs. immediate sex)

28
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.

29
VARIABLE RATIO - reinforcement after a random
of responses. Greater resistance to extinction,
but initial learning is slower than continuous
reinforcement. Ex Gambling.
REINFORCEMENT SCHEDULES
FIXED INTERVAL -reinforcement after a specific
amount of time. Ex Checking the mail.
FIXED RATIO-reinforcement after specific of
responses. Ex Piece-work
VARIABLE INTERVAL-reinforcement after random
amount of time. Most resistant to extinction. Ex
Pop Quizzes
30
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 is always reinforced gives evidence of
learning throughout trials.
Rat that explored the maze for 10 days, then
given a reward on Day 11, shows signs of Latent
Learning.
31
Hey kids! Ill pay you to play with that toy
Tang Hall Study (1995)
This toy is great-Im going to play with it all
the time!
If Im not getting paid, why bother. OR If I
have to be bribed, then it must not be that great
of a toy.
32
OVERJUSTIFICATION EFFECT-decreasing intrinsic
motivation for a behavior after an extrinsic
reward is provided.
Salary
Performance
33
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.

34
Skinners Legacy
  • Give rewards that promote desired behavior
  • Personal freedom? Not in Skinners world.

35
Operant Conditioning Applications
School Work Home
Computers 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
36
Albert Banduras Bobo-doll experiment Aggressive
behaviors learned through observation of a model.
37
(No Transcript)
Write a Comment
User Comments (0)
About PowerShow.com