Learning

Learning

Learning is a relatively permanent change in behavior that occurs through experience.

How Do We Learn?

• learning that certain events occur together

Classical Condition

• The type of learning in which a response naturally elicited by one stimulus becomes to be elicited by different formally neutral stimulus
• Pavlov and Watson
• USC, UCR, CS, CR

Operant Conditioning

• The type of learning in which behaviors are emitted to earn rewards or avoid punishment 
• B.F. Skinner
• Reinforcement and punishment

Social Cognitive Learning Theory

• The type of learning in which behaviors are learned by observing a model
• Albert Bandura 
• modeling and vicarious learning
• an involuntary behavior is determined by what precedes it 

Unconditioned Stimulus (US): A stimulus that naturally and automatically triggers a response

Unconditioned Response (UR): The unlearned, naturally occurring response to the UCS

Classical Conditioning: neutral stimulus (something that by itself elicits no response)

Conditioned Response: originally neutral stimulus (NS) that after, association with the UCS, comes to trigger a response

Conditioned Response (CR): the learned response to a previous neutral stimulus

- the responses (UR & CR) are always the same

- The NS and CS are always the same

- The NS becomes the CS through learning

5 Critical Principles Are:

Acquisition

• we know learning exists because the CS is linked to the USS
• the CS should come before the USC
• they should be very close together in timing 
• learning is better when the UCS and CS occur together frequently
• learning is better when the CS is new or unique

Extinction 

• acquisition doesn't last forever
• the moment the CS is no longer associated with the USC, we have extinction
• Reconditioning: the quick relearning of a CR after extinction 

Spontaneous Recovery

• sometimes, after extinction, the CR still randomly appears after the CS is presented

Generalization

• sometimes is so similar to the CS that you get a CR 

Discrimination

• something so different to the CS you do not get a CR

Operant Conditioning

• The learner is NOT passive
• learning bases on consequences
• A type of learning in which behavior is strengthened, it is followed by a punishment

Classical vs. Operant 

Classical:

• behavior is determined by what PRECEDES it
• involuntary
• dog salivates after a tone

Operant

• behavior is determined by anticipation of what FOLLOWS it
• voluntary
• dog sits in anticipation of getting a treat

They both use acquisition, discrimination, spontaneous recovery, generalization, and extinction

Law and Effect

• Thorndike's principle the responses are "stamped in" by rewards and "stamped out" by punishments

Shaping

• is reinforcing small steps on the way to the desired behavior

Reinforcer

• any event that strengthens the behavior it follows
• 2 types: positive and negative

Positive Reinforcement strengthens a response by adding a stimulus after a response

Negative Reinforcement removes something unpleasant that was already in the environment following a behavior, making that behavior more likely to occur again (think avoiding/escaping)

Punishment

Any event that DECREASES the behavior that it follows 

Meant to decrease behavior

Positive Punishment:

• addition of something unpleasant

Negative Punishment:

• removal of something unpleasant

Punishment works best when it's immediately done after behavior and if it harsh

Types of Reinforcers 

Primary Reinforcer

• An innately reinforcing stimulus

Secondary Reinforcer

• A stimulus that gains its reinforcing power through its association with a primary reinforcer

Token Economy

• every time a desired behavior is performed, a token is given
• they can trad token for a variety of prizes (reinforcers)
• used in homes, prisons, mental institutions, and schools

Continuous vs. Partial Reinforcement

Continuous:

• Reinforce the behavior EVERY TIME the behavior is exhibited
• usually done when the subjects is first learning to make the association
• acquisition comes really fast
• but so does extinction

Partial:

• Reinforce the behavior only SOME of the times it is exhibited
• acquisition comes more slowly
• but more resistant to extinction
• FOUR types of partial reinforcement schedules

Fixed Ratios 

- provides a reinforcement after a set number of responses 

Variable Ratios

- provides a reinforcement after a random number of responses

- is very hard to get acquisition but also very resistant to extinction

Fixed Interval

- requires a SET amount of time to e;apse before giving the reinforcement

Variable Interval

- requires a random amount of time to elapse before giving the reinforcement

- very hard to get acquisition also very resistant to extinction

Observational Learning

• Albert Bandura and his BOBO doll
• we learn through modeling behavior form others
• observational learning + operant conditioning = social learning theory

Latent Learning

• Edward Toleman
• latent meas hidden
• sometimes learning is not immediately evident

Insight Learning

• some animals learn through the "ah ha" experience