Positive Reinforcement training questions - German Shepherd Dog Forums

Increase font size: 0, 10, 25, 50%

GermanShepherds.com is the premier German Shepherd Forum on the internet. Registered Users do not see the above ads.
Reply
 
LinkBack Thread Tools
Old 01-22-2014, 09:39 AM   #1 (permalink)
Member
 
dpc134's Avatar
 
Join Date: Jan 2013
Location: Pennsylvania
Posts: 150
Default Positive Reinforcement training questions

All,
I have been doing some research on Positive Reinforcement training methods and I am curious on a few scenarios how this method works for certain things. I trained my dog certain commands (sit, down, stay, etc.) using +R only. But how does +R work to teach unwanted behaviors?

How would you train your dog to stop chewing on something using Positive Reinforcement only?
How would you use +R to stop your dog from play biting too hard with another dog?
What about high distraction situations, when the distraction (another dog, person, etc.) is more motivating to your dog than your treat? How do you handle those situations?

I am only asking because I am trying to learn and expand my training knowledge. I have successfully trained my dog in all of these situations, however, I have not used +R only. I am curious how +R would be used to train these situations.

Thanks in advance.
dpc134 is offline   Reply With Quote Quick reply to this message
Sponsored Links
Advertisement
 
Old 01-22-2014, 10:11 AM   #2 (permalink)
Master Member
 
Sp00ks's Avatar
 
Join Date: Nov 2013
Location: NC
Posts: 824
Default

Subscribing to this one as I'm curious about the scenarios you have outlined as well. I was about to contract my trainer with nearly these exact questions. (It's been 10 years since we last trained a pup)

I have seen people talk about +R and use prong collars or E collars. I'd like an explanation on that as well. It just doesn't seem like +R. I'm not saying I don't agree with some of the training tools but I don't see how they can be claimed as +R.

In our case, a High distraction could be a stick or blowing leaf at this point
__________________
Dagr "Dag" 11/07/2013-
Sp00ks is offline   Reply With Quote Quick reply to this message
Old 01-22-2014, 10:48 AM   #3 (permalink)
Member
 
dpc134's Avatar
 
Join Date: Jan 2013
Location: Pennsylvania
Posts: 150
Default

I hear ya. I would love to train +R only. Prong and E-collars are definetly not +R. I currently use Prong and E-collar and I do not enjoy using corrections with my dog, but it has worked very well.
I trying to understand how to use +R only to train this behaviors / situations instead of a Prong and E-collar. People always mention +R only and I am curious how they would handle this scenarios.
dpc134 is offline   Reply With Quote Quick reply to this message
Old 01-22-2014, 12:03 PM   #4 (permalink)
Elite Member
 
Join Date: Mar 2006
Posts: 1,610
Default

I think for most +R "only" exists only in the mind

To take one scenario, dogs or other things being more highly prized than you and what to do about it.

First, you really need to know your dog and what it is ready for, next is managing that. You have to keep them from self rewarding on something that is higher value than you at that moment. I always use a drag line I can step on.

Not completely +R because if they try and leave, they do hit the end of the leash. It's more teaching them that they can't just check out and go do what they want and when they get frustrated, they come back and find great stuff still comes from me. That happens very infrequently because I don't put my dogs in that position.

You always have to be mindful of what is going on around you and how competing the distractions are. If you build it right, you don't have a dog that just checks out and goes for whatever it wants, but in the real world with real people and real dogs, you don't always get it "right" and there will be times you're in a situation that will call for something else.

Usually just by limiting what a dog can do during "learning" and building up to more and more competing distractions in a controlled environment, you can eliminate so much of what others struggle with all the time in training.

Chewing on things? that's a management thing. puppies chew, give them appropriate things, remove other things and limit their options.

I don't know how you'd control play biting, that's not something I think is worth even trying to think thru how I'd simulate that and transfer it to dog/dog interactions. I can certainly stop play at any point by giving them another command i've taught thru +R interactions and then let them resume when they've calmed a bit. But I don't think that's what you're asking.

+R only usually means +R "mostly" in the real world
crackem is offline   Reply With Quote Quick reply to this message
Old 01-22-2014, 12:10 PM   #5 (permalink)
Moderator
 
Liesje's Avatar
 
Join Date: Mar 2007
Location: Grand Rapids, MI
Posts: 18,542
Default

Remember what +R really means. You can't really talk about one quadrant of operant conditioning without understanding all four. "Reinforcement" means the behavior happens more frequently; "punishment" means the behavior happens less frequency, to the point of "extinction". So based on the actual theory behind the terms, you *can't* use +R to stop a dog from chewing shoes, etc. What you CAN do is simply manage the environment and/or reinforce incompatible behaviors. I can "reinforce" my puppy chewing a Nylabone and playing tug on a tug toy without having to "punish" him for chewing shoes.
__________________
Coke (All-American 7/7/06)
Nikon (GSD 9/7/08)
Indy (All-American 5/10/12)
Legend (GSD 10/22/13)
Rainbow Bridge Kenya (GSD)

Liesje is offline   Reply With Quote Quick reply to this message
Old 01-22-2014, 12:15 PM   #6 (permalink)
Master Member
 
Join Date: Jan 2013
Posts: 875
Default

Quote:
How would you train your dog to stop chewing on something using Positive Reinforcement only?
DW posted this vid in another thread.

Pretty interesting technique and perspective on training an off command with positive only methods.

https://www.youtube.com/watch?v=ndTiVOCNY4M

A similar theory is to offer a new toy or ball or tug in return for dropping another.

Quote:
What about high distraction situations, when the distraction (another dog, person, etc.) is more motivating to your dog than your treat?...I am curious how +R would be used to train these situations.
The environment needs to be controlled and the dogs obedience or compliance needs to be built up introducing distractions over time.

Also if you really build drive for a toy, the dogs desire to get it out weighs it's interest in distractions.

But still that is the theory, if your dog is staring at a cat and ready to chase a little negative reinforcement is needed.
MadLab is offline   Reply With Quote Quick reply to this message
Old 01-22-2014, 12:16 PM   #7 (permalink)
Master Member
 
Sp00ks's Avatar
 
Join Date: Nov 2013
Location: NC
Posts: 824
Default

Quote:
I can certainly stop play at any point by giving them another command i've taught thru +R interactions and then let them resume when they've calmed a bit. But I don't think that's what you're asking.
Personally, I have started doing just that. Too soon to tell how well it's working. I know he is amped up and doesn't want to do anything but continue playing. However, I make him sit until I can let him go and he remains sitting while the distraction is still present. When I can tell his heart rate and breathing have dropped, I'll let him resume play. "Free"
__________________
Dagr "Dag" 11/07/2013-
Sp00ks is offline   Reply With Quote Quick reply to this message
Old 01-22-2014, 12:19 PM   #8 (permalink)
Knighted Member
 
Baillif's Avatar
 
Join Date: Jun 2013
Location: North Carolina
Posts: 2,480
Default

+R only is pretty impossible anyway. Anytime the dog doesn't perform a behavior or performs the wrong one and you withhold reward you're technically doing -P at that point.

I'll take a shot at it though.

In the case of getting the dog to stop mouthing something you teach the off command with the treat in the hand you want the dog to stop mouthing. There is a tab dog training video on YouTube if you need to see an example. When they stop messing with it you mark reward. You can generalize that with other objects. I know because I have.

In the case of the last two you don't use treats as the reward. You use what the dog wants in those scenarios as the reward. You wait until they do whatever you want from them before allowing them to do what they want to do. This of course assumes you can actually do that. They might want to hump some poor little old lady's leg. If you go this route you might need to set up scenarios to teach it. As before there are elements of negative punishment in there but I don't really think those are avoidable.

Liesje is technically correct though. You aren't stopping a behavior in those cases you are just creating an alternative incompatible behavior the dog could choose over the behavior you don't want.

Last edited by Baillif; 01-22-2014 at 12:22 PM.
Baillif is offline   Reply With Quote Quick reply to this message
Old 01-22-2014, 12:24 PM   #9 (permalink)
Knighted Member
 
wolfy dog's Avatar
 
Join Date: Aug 2012
Location: USA
Posts: 2,732
Default

+R doesn't mean you should never punish your dog. Punishment in these newer techniques consist of withholding the payoff the dog is seeking. For instance: if your dog tries to bolt out the door as soon as you open it and you close the door in front him, he is punished and learns that when he acts this way, he is not going anywhere. Punishment doesn't have to equal choking, applying electric shocks, hitting, yelling or anything else people come up with. And yes, controlling the environment makes your job a whole lot easier. If a pup never has had the opportunity to chew on shoes he will not consider them toys when he is grown up.
wolfy dog is offline   Reply With Quote Quick reply to this message
Old 01-22-2014, 12:29 PM   #10 (permalink)
Moderator
 
Liesje's Avatar
 
Join Date: Mar 2007
Location: Grand Rapids, MI
Posts: 18,542
Default

Quote:
Originally Posted by wolfy dog View Post
+R doesn't mean you should never punish your dog. Punishment in these newer techniques consist of withholding the payoff the dog is seeking. For instance: if your dog tries to bolt out the door as soon as you open it and you close the door in front him, he is punished and learns that when he acts this way, he is not going anywhere. Punishment doesn't have to equal choking, applying electric shocks, hitting, yelling or anything else people come up with. And yes, controlling the environment makes your job a whole lot easier. If a pup never has had the opportunity to chew on shoes he will not consider them toys when he is grown up.
I agree but people should be clear that this is not "positive reinforcement". -P is -P, not +R. I agree with previous posters that training a dog completely +R is basically impossible. There is also classical conditioning at play as well.
__________________
Coke (All-American 7/7/06)
Nikon (GSD 9/7/08)
Indy (All-American 5/10/12)
Legend (GSD 10/22/13)
Rainbow Bridge Kenya (GSD)

Liesje is offline   Reply With Quote Quick reply to this message
Sponsored Links
Advertisement
 
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the German Shepherd Dog Forums forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in

Human Verification

In order to verify that you are a human and not a spam bot, please enter the answer into the following box below based on the instructions contained in the graphic.


Thread Tools

Posting Rules
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On



All times are GMT -4. The time now is 01:54 PM.



Powered by vBulletin® Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
SEO by vBSEO 3.3.2
PetGuide.com
Basset.net DobermanTalk.com GoldenRetrieverForum.com OurBeagleWorld.com
BoxerForums.com DogForums.com GoPitbull.com PoodleForum.com
BulldogBreeds.com FishForums.com HavaneseForum.com SpoiledMaltese.com
CatForum.com GermanShepherds.com Labradoodle-dogs.net YorkieForum.com
Chihuahua-People.com RetrieverBreeds.com