Why don’t you shoot the dog ?
The book is about animal training, behavior modification, not really killing any animal.
To start with , reinforcement is better than results. A reinforer is anything occur in conjunction with an act, tend to increase the probability that the act will occur again. There are 2 type of reinforcer: positive / negative.
Behavior that occurring, can always be intensified with positive reinforcement. A major point in training with reinforcement is you cant reinforce behavior that is not occurring.
In order to be re-inforcing, item hose must be something subject want. Necessity of switching constantly from to another is challenging and interesting.
Negative is not same as punishment. Punishment occur after the behavior it was meant to modify and therefore no effect on the behavior.
Use of retreat, easing back when desired behavior occur, is an important aspects of most called ” horse whisper” technique.
Timing of reinforcer, timing of arrival of reinforcer is information. The informed content of a reinforcer become more important than the reinforcer. First question on training problem is “Is the reinforcer too lat”. Size of reinforcer, to start with , as small as you can get away with. Smaller the reinforcer, faster finish more reinforcer per session.
One extremely useful technique is the JACKPOT. 10 times bigger than normal reinforcer.
Conditioned reinforcer – initially meaningless signal, presented before/during delivery of reinforcer. Once you have conditioned reinforcer, be careful not to throw it around or it will dilute its force. A timely conditioned positive reinforcer, what you doing now is good and will gain you something. Conditioned Aversive, what you do now is no good, something bad will happen unless you stop. Re-primed are necessary part of existence.
Constant reinforcement is needed just in learning stages. To maintain a learned behavior, we reinforce on a random/UN-predictable basis. Overcome slow start phenomenon is to introduce some re-reinforcement just for getting started.
Group recognition are powerful reinforcer. Shaping consist of taking a small tendency in right direction and shifting it, by successive approximation. Success / failure in shaping a behavior depends on our persistence. A well planned shaping program can minimize the required drilling and make every moment counts. 2 aspect to shaping : 1. Method : the behavior to be developed and the sequence of step. 2. Principle : Rule governing how / when/ why theses behavior are reinforced.
10 General rules for shaping
- Raise criteria increments small enough that the subject always has a realistic chance of re-reinforcement. How fast you raise the criteria is not a function of subject’s actual ability. Its a function of how well you are communicating through – your shaping procedure, what your rules are for gaining reinforcement. Every time you raise a criteria, rules is changed, subject need to give the chances to discover it. Constant progress will get your ultimate goal faster than try to force rapid progress, risk of losing good performance.
- Try to aim for train 1 aspect of behavior a a time, don try to shape for 2 criteria simultaneously. If task can be broken down into separate component, then shape separately, learning will go much faster. When we see no progress in a skill, normally its because we try to improve 2 or more things at once. Practice is not shaping. Repetition may in grain mistake as easy as improvement.
- During shaping, put current level of response on to a variable schedule of reinforcement before adding/raising criteria. Sometimes you reinforce, sometimes you don’t. When we teach the behavior – put in a fixed schedule of reinforcer, when we want to monitor the behavior – variable schedule.
- When introducing a new criteria, aspect behavior skill, we could temporary relax the old one. Once learning is not forgotten, but under pressure of assimilation new skill level, old well-learned behavior sometimes fall temporary
- Stay ahead of your subject. Plan your shaping program so if your subject make a sudden leap forward. You know what to reinforce next. A breakthrough is a golden opportunities to make many progress in hurry.
- Don’t change trainer in midstream
- If one shaping procedure is not eliciting progress, try another. It’s important to understand the “principle” and not just recipes. The principle govern what truly work.
- Don’t interrupt a training session roughly, that composite a punishment. When a subject easy to earn reinforcer, its a contract with the trainer. Removal attention is a good way to re-balance a subject.
- If a learned behavior deteriorates, review the shaping. Quickest way to connect deterioration is to get the subject recall the original shaping procedure and go all the way through it rapidly. “Go back to kindergarten.”
- Quit while you are ahead. When is stop is not as important as what you stop. Always quit while you are ahead. Always move on a high note.
Shaping short cuts
Targeting : Shape an animal to touch a target, then move the target around, for example the flag /banner in tourist group
Mimicry : watch and copy. learn by observation. To teach physical skill to human, mimic is a good technique.
Modeling : Push the subject manually through action we want them to learn. Put a subject through the same motion long enough and often enough, it would learn how to do the behavior.
Self-shaping : Record Keeping, record performance in such a way improvement can be seen at a glance , for example on a graph.
Anything can cause behavior response is a stimulus. There are 3 way to add cue 1. conditioned negative reinforcer 2. Conditioned positive reinforcer 3. Shape response to the cue.
Reward response to even one learned command, people rapidly start responding to other command to earn reinforcement.
There are 8 methods to get rid of behavior you don’t want
- Shoot the animal
- punishment – it normally doesn’t work as it occur after the behavior. Side effect of punishment – hate /fear / anger to the punisher.
- negative reinforcement – introduce unpleasant event / stimulus, stop or avoided by changing one’s behavior.
- Extinction – Behavior die down by itself for lack of re-forcement. Behavior provide no result (good / bad) will probably extinguish. Ignoring the behavior within ignoring the person.
- Train an incompatible behavior – Perform another behavior physically incompatible with the one you don’t want.
- Put behavior on cue – Do the crazy version of behavior in a given time
- shape the absence – Good when you want to stop what he is doing, but don’t have anything particular wish him to do
- Change the motivation – Eliminate the cause, or change the motivation. Full people don’t steal bread.
Good book to introduce how to train our-self / others people / your neighbor’s dog ,until next time
Leave a Reply