Direktlänk till inlägg 28 november 2011
På engelska, jag klara inte att lyssna på engelska och anteckna på svenska...Och som vanligt är det mina personliga anteckningar som jag hoppas är förståeliga också för någon som inte är jag.
Ken Ramirez: Become the best trainer you can be
General aspects of training: species does not matter. Training is the same – we all learn the same way! Of course species are different, and you need to know that, but the training is the same.
Goals for the day:
1.Training philosophy of Ken
2.Cross-species training: what can I learn from zoological training
3.Dealing with mistakes in a positive way (mistake an animal make)
4.Advanced tools
1.
What is training?
Definition: Training = teaching! Teach the animal how to live in our world. Teaching is a shared process: we respond to the animals, the animals respond to us. It is a natural process, the mother teaches the children, animals always learn! Our responsibility is to teach them how to have a good life in OUR world.
Cornerstones of animal care:
1.Health care (veterinary program)
2.Nutrition (food and vitamins)
3.Environment (includes social structure) Ex: cats like to scratch and climb, get them cat-trees indoors. Horses are made to live together, allow them to do that.
4.Behaviour management (training and enrichmenmt)
All 4 are necessary! Training is JUST AS IMPORTANT as the others.
Primary reasons for training:
Anything that benefits the individual.
1.Physical exercise
2.Mental stimulation
3.Cooperative behavior
Secondary reasons for training:
Education, Research, Conservation, Entertainment, Work animals, Sport, etc…
As a trainer my first responsibility is towards the INDIVIDUAL! The animal needs must come first! Get your priorities in order.
The rules of learning is easy, but behavior is complex.
2.
Cross species training. How hard is it? How critical is it to understand the science? When do differences impact training?
Everything starts with GETTING TO KNOW THE ANIMAL!
Species:
Easier to start training a natural behavior and get in on cue. When the animal understands the training game you can teach complex things.
Individual:
Knowing the animal
Trust and relationship building
Determine type of contact. Fences and protection? Six types of contact:
1.Free contact
2.Semi protected (muzzle, leash)
3.Protected contact (other side of a fence or other barrier). The animal chooses to stay or leave, you can still touch some part of the animal through a window (ear, snout, paw, etc.)
4.Confined contact (the animal can not leave)
5.Hands-off training (fences or barriers, no physical contact) You can train without touching!
6.Remote training – designed to NOT build relationships, no human involved. Prepare for reintroduction in the wild, or create an interest for the environment. Reinforcement will come from the environment, by remote control or similar.
Developing a relationship.
Animals learn 24 hours a day! USE that! PAY ATTENTION!
Getting started (trainingprogram)
1.Choosing the right reinforcers
2.Select an event marker – audible, tactile, visual. Not necessary, but helpful!
3.Operant or classical conditioning?
4.Pick the right place to begin. Goal? Starting point for behaviour?
Case study 1: targeting
Exemples: Nosetouch, targetstick, pawtarget, glove/perch for bird, mat for dog, freezetarget, follow target, grab target, experienced sleddogs is a target for youngsters, stretch out paw through window, laserpointer
8 basic target skills:
1.Basic target (first we touch mark and feed, then we wait for the animal to touch and we mark and feed)
2.Different types of targets (stick, plate, hand etc)
3.Complex (use the target to teach something else, let the animal reach further, search for)
4.Extended (freeze)
5.Follow
6.Different bodyparts
7.Simultanious targets (nose AND paw)
8.A to B (move from trainer to trainer, two locations)
Case study 2: kenneling (cage, box, through a doorway or over a threshold)
1.Work with a partial kennel, build it bit by bit.
2.Give treats inside the kennel, build a high value
3.It is a build-up of trust
4.Keep the balance of reinforcement so that they want to leave the kennel…
Case study 3: tactile
Ex 1: Owl – terrified of people. 2,5 years to be able to touch its head, back, talons, etc. Gain the trust, TRAIN! 5 years until everybody on staff was able to touch it.
Ex 2: Too tactile…otters trained to grab something else, a target, the wall.
Ex 3: Tapir, frightened. Desensitization all over the body, the tactile became reinforcing! Massage was very effective.
Handling animal mistakes – positively
Why is this important? Which are the tools and techniques? Popular but aversive techniques under the microscope. Positive alternatives! Using redirection - how?
Traditional approach: punish. Punishment works. The very definition of a punisher is that it makes behavior GO AWAY. Why do we want to avoid them? They lead to fear, aggression, and passive animals. But is positive reinforcement as effective? And which are the best PR techniques?
“Don´t shoot the dog” is all about getting rid of unwanted behavior. 8 different techniques. Read it!
Changing motivation is the long term solution. It is not always possible in the moment. This lecture is about here and now.
Positive reinforcement training:
What does that mean? Aversives are NEVER present? Hierarchy of effective procedure (“list of procedures”) by Dr susan Freedman:
1.Medical, nutritional, physical reasons must be examined first of all
2.Antecedant arrangements. Set the environment up for success. Take the temptation away.
3.Positive Reinforcement training
4.Alternative behaviours – redirection
5.Extinction, negative reinforcement, negitive punishment
6.Positive punishment
If you TRY and fail all the steps before you use punishment, you are a Positive Trainer (according to Ken). You should take every opportunity to become a better trainer! Often the reason for punishment is that the trainer is not as skilled in reinforcement methods as she/he could be, or should be.
This lecture will focus on 6, 5, 4.
6.No reinforcement Marker (felsignal) is a positive punisher! We want a behavior to end, and we bring something to the situation. We mark the moment the animal does a wrong move. The opposite of a clicker, a conditioned punisher. If you use it WELL, the animal should be HAPPY and NOT FRUSTRATED – but it takes skill!
If it works for you – use it. But don´t overuse it! We tend to fall into a habit of “no”. Selfdiscipline! Pick a special word, that you don´t use every day, to really have to make the choice to use it (no more “I said no by accident, I´m so used to it”)
Deltasignal: “last chance”, warning “if you don´t change behavior I will not reward you. If you correct yourself I will.” Alternative: don´t reinforce that time, try again later – a little easier.
Not now, keep going, good try: are they necessary – or just superstitious for the trainer? Will it really speed up the training?
“Try harder-signal”: high energy dog will be pushed harder and be more motivated. BUT it takes a certain dog and a really skilled trainer. Analogy from Ken´s life: he was pushed very hard by his trainer when he trained as an athlete, and took the verbal abuse as “I´m going to do better!” It can work, but is it necessary? Would he have gotten the same results with another – not abusive – trainer? We don´t know.
Delta-signal often becomes a cue. Or even THE cue. Delta signal = “…or else”-cue.
5.Time-out
“Removal of opportunity for reinforcement” Negative punisher.
Animal exhibits undesired behavior, trainer leaves or turn back to animal. Many trainers use them TOO MUCH! Frequently the animal gets frustrated and does not understand. Timing is everything!
Time-outs are effective – in the right conditions! And those conditions are rare. It can in worst cases ruin your relationship with the animal.
Ken avoids – except if the ANIMAL gives YOU a time-out (leaves).
Taking a break for my frustration or my animals frustration is NOT a time-out. It´s a break. End with something simple and easy.
5.Negative reinforcement
Removal or avoidance of an aversive stimulus will increase the frequency of a behavior.
Exemple: horse and reins. Positive punisher for “going straight”, negative reinforcer for turning. Becomes a cue. Properly used it WORKS! Alarmclock: it rings until we get up, our getting up is negatively reinforced. Yes, we use an aversive, and have to think about what does the dog percieve the most? The punish or the reinforce? NR happen in our world all the time! Can be a very humane and effective choice, if done correctly.
Ex: following collar pressure. The aversive is VERY mild.
Ken prefers negative reinforcement to negative punishment, because your focus is still on getting behavior and not stopping it.
Reinforcement: focus on GETTING behavior – help do the right thing. Punish: focus on STOPPING behavior – punishing the animal.
Overuse can cause frustration. Severity of aversives are hard to control for many. Inappropriate use can be inhumane. Skilled trainers only!
5.LRS
Least Reinforcing Stimulus/scenario.
Designed as most positive approach to incorrect behavior. A way to ignore unwanted behavior that cause minimum of frustration.
3-5 second neutral response. Nothing happens.
Goals:
Practical use:
Challenges:
Why does it work:
Extinction takes time! You can get extinction bursts, but it is rare with this technique.
Can be used in a shaping session, always follow up with an easy cue. Be aware of patterns! Combine with redirectional techniques.
Default position or station to return to after LRS helps to “reset” the animal
4.DRA
Alternate response training: Differential Reinforcement of…Incompatible behavior, Other behavior, Lower intensity behavior. Show the animal what you WANT it to do!
DRI:
Ex: Hold on to a target instead of grabbing trainer. Teach a new “begging place”
DRO:
Ex: Baby elephant in the way gets reinforced for everything that is not “walk next to mummy”, Dolphins get the toys when they do anything that is not aggressive
DRL:
A difficult technique to master, we are still clicking the undesired behavior.
Simplifying complex training tools
Exploring advanced concepts:
Science vs practical application
Overview of advanced tools:
Chained behaviours
Technical or common? Technical: Each behaviour starts the next and reinforces the former. Common chain: sequence of behaviors, bridge after every behaviour, cues are used.
Building a chain:
Train the animal to accept new reinforcers. In a backchain we want the nest behaviour to be the reinforcer. This makes a chain stronger.
Most behaviours are small chains. We build duration and length through approximation.
A well trained chain is NOT a variable schedule, but a reinforcement variety.
Jag uppskattade speciellt hans tankar om hur man tacklar att djuret gör fel. Ken är alltid noga med att betona att detta är HANS åsikter, och vi får gärna tycka och tänka annorlunda. Jag fick mig definitivt en tankeställare gällande felsignaler och negativ förstärkning, nya tankar om hur jag kan använda LRS, och som alltid inspiration och motivation! Flera frågor återstår, men jag hann inte stanna och fråga efteråt (det tog över fem timmar att köra hem i stormen...). Det blir till att formulera ett mail senare i veckan!
Må | Ti | On | To | Fr | Lö | Sö | |||
1 |
2 | 3 |
4 |
5 |
6 | ||||
7 | 8 |
9 |
10 |
11 |
12 |
13 |
|||
14 |
15 |
16 |
17 | 18 |
19 | 20 |
|||
21 |
22 | 23 | 24 |
25 |
26 | 27 |
|||
28 | 29 | 30 |
|||||||
|