13 Animal Training Flashcards
How animals (including us) learn (3)
- Innate (instinct and habituation)
- Latent/Exploratory Learning
- Associative learning (conditioning)
Instinct learning
usually related to hunting, finding a mate or territorial behaviour; under normal circumstances, the learning occurs during play with littermates
-tracking scent, catching mice
latent/exploratory learning
involves making associations without immediate reinforcement/reward
- allows it to learn about it’s surrounding as it explores (door, bowl)
Conditioning (2)
- classical conditioning (learning to recognize)
2. Operant conditioning (learning to respond)
Classical conditioning and example
associates an involuntary response and a stimuli (creates meaning in an unconditioned stimulus; bridging)
example: clicker training; click=treat, eventually the sound of click may become rewarding (need perfect timing)
Operant training
forms association between a behavior and a consequence; focuses on either using reinforcement/punishment to increase/decrease a behavior
* consequences have to be immediate/ clearly linked to the behavior
Operant training: consequences to any behavior (4)
- something good can start/be presented
- something good can end/taken away
- something bad can start/be presented
- something bad can end/be taken away
superstitious behaviors
common in horses and dogs, where if they do one thing and do smth else and get rewarded for the first thing, they might think it’s the combination of both that got them the reward
shaping and example
specific behaviors are more likely to exist on a gradual, continuous scale; must gradually shape behavior until ideal form is displayed
eg if want cat on the mat, start by rewarding when cat even looks at the mat; gradually work towards it (one step, two steps)
dont give too much rewards in the early stage but give jackpot in the end
shaping: guidelines
- only reward behavior if it’s closer to ideal than previous
- always look for improvement; too much reward may fix less than ideal behavioural form
- lowest step will be the simplest version or the very first piece of it; reinforce successive approximations
- the moment they understand it, do it more times then level up
Positive vs negative
+: something added
-: something taken away,
Reinforcement vs punishment
R: anything to increase behavior
P: anything to decrease a behavior
Training examples
- +Reinforcement: dog gets a treat sitting
- +Punishment: dog squirted with citronella when barks
- -Reinforcement: trainer takes off pressure once dog lays down
- -Punishment: dog put on a leash and taken from the park for coming to the owner when called
* Not all positive is good and not all negative is bad
Motivation for training
- what motivates one animal will not motivate another; trainers need to understand what the motivators are for the particular animal they are working with
- most trainers expect the animals to fit in with the motivation they use (eg tug toy) and dismiss an animal thats not motivated to work with their motivator
Caution for motivation
- should be seen as payment for a job done
- each animal has a different pay scale, some requires more than others; up to the trainer to know what the individual animals pay scale is
- how much you give will keep their motivation going but must strike balance and keep them interested and not full (wont be interested anymore)
Types of motivation/reward (4)
- toys
- food
- game, body contact (pat)
- relaxing in body posture (removing stimulus/relief)
* may also depend on daily preference, switch it up
Timing and ratio of rewards to behavior and example
the more irregular the pattern of the reward, the harder it is to extinguish the behavior
eg an animal used to getting a reward on a variable ratio/interval will carry on performing the behavior compared to one used to frequent reward
Why is variable rate/ratio better?
doing the behavior becomes addictive; if they keep doing it, reward will come eventually
Intrinsic rewards
-those that originate from within the dog; related to feeling good (usually)
-often stronger than extrinsic
eg, happy feeling when postman drives off and he barks, excitement of chasing bird
Downside of intrinsic rewards
difficult to get rid of, eg chasing a bird) bc the action itself is rewarding (even more rewarding than treats)
-punishing them for something self rewarding will not work
Extrinsic rewards
-those that originate form something beyond the dog; physically given for performing an action or behavior
eg giving the dog a treat for sitting, removing lead after dog has pulled you to the park
Breed specific training and reward (3)
- dog that has been bred to herd (collies) may prefer to chase a ball rather than a game of tug
- lap (spaniel) might prefer food/pats to playing with ball
- dog that is bred for biting and killing (terrier) might prefer a game of tug to food
Why is breed specific training important?
problems arise when people dont understand the needs of their dogs
- people often try to mould a dog to needs but instinct will usually override training
Reward vs motivation
Motivation comes before the action; what drives the dog to take action or perform the behavior
Reward comes after the action; what the dog gets from taking the action or performing the behavior
Dog motivators are: (4)
- breed specific
- age specific (most get more food motivated when they get older)
- environment specific
- past experience may affect ability to learn
Appetitive vs aversive stimuli
- appetitive stimuli is something the dog finds pleasing (food, toys, play body contact)
- aversive stimulus is something the dog finds unpleasing (collar jerk, social isolation)
Rewards in terms of appetitive and aversive stimuli
a reward can be purely appetitive or can be relief from something the dog finds aversive
importance of engagement
engagement: dog WANTS to be with you and wants what you have (toy, food)
- if dont have engagement, how to train?
- when you try to force something, you will always be met with resistance
