diff --git a/yr/images/.DS_Store b/yr/images/.DS_Store index 8d04a21..7c55370 100644 Binary files a/yr/images/.DS_Store and b/yr/images/.DS_Store differ diff --git a/yr/images/RL_tamethedog/BlocksEditor.png b/yr/images/RL_tamethedog/BlocksEditor.png new file mode 100644 index 0000000..4c2c86f Binary files /dev/null and b/yr/images/RL_tamethedog/BlocksEditor.png differ diff --git a/yr/images/RL_tamethedog/TabletSize.png b/yr/images/RL_tamethedog/TabletSize.png new file mode 100644 index 0000000..9c8ffd0 Binary files /dev/null and b/yr/images/RL_tamethedog/TabletSize.png differ diff --git a/yr/images/RL_tamethedog/WhenResetButtonClick.png b/yr/images/RL_tamethedog/WhenResetButtonClick.png index 069e47d..230f6e2 100644 Binary files a/yr/images/RL_tamethedog/WhenResetButtonClick.png and b/yr/images/RL_tamethedog/WhenResetButtonClick.png differ diff --git a/yr/tutorials/RL_TametheDog.md b/yr/tutorials/RL_TametheDog.md index ff429af..9574e0f 100644 --- a/yr/tutorials/RL_TametheDog.md +++ b/yr/tutorials/RL_TametheDog.md @@ -50,9 +50,17 @@ Where Q(c,a):Q[command,action] and α: Learning Rate To allow you to focus on the new ideas of RL in this project, the User Interface (UI) has been created for you. ![GUI explained](../images/RL_tamethedog/GUIExplained.png){:.enlargeImage} +Note: If the entire UI is not visible in the default Phone size given in the Viewer, try switching to the Tablet size to be able to see everything together. + +Tablet Size + ## Variable Initializations +Now switch to the Blocks editor. + +![Blocks editor](../images/RL_tamethedog/BlocksEditor.png){:.enlargeImage} + The coding starts by initializing the variables you will use in the programming. ![Variable initializations](../images/RL_tamethedog/Initialize.png){:.enlargeImage} @@ -180,7 +188,7 @@ In this code, see if you can explain what happens when the Reset button is click -Qtable entries are re-initialized to 0’s, and these values are displayed via the printQTable procedure on the UI. The variables counter and reward are set to 0, and the variables command and action are also initialized to STAND (or the number 0). learningRate is re-initialized to 0.5, as is the LearningRateSlider.ThumbPosition. The dog’s image is set to the “standing” image. RewardButton and PenaltyButton are both disabled, and the SitButton and ShakeButton are enabled. Everything resets to what it was before the training started. +Qtable entries are re-initialized to 0’s, and these values are displayed via the printQTable procedure on the UI. The variables counter and reward are set to 0, and the variables command and action are also initialized to STAND (or the number 0). learningRate and its label are re-initialized to 0.5, as is the LearningRateSlider.ThumbPosition. The dog’s image is set to the “standing” image. RewardButton and PenaltyButton are both disabled, and the SitButton and ShakeButton are enabled. Everything resets to what it was before the training started.
@@ -314,7 +322,9 @@ Now test your app by scanning the QR Code generated via your AI2 Companion. ![Scan QR Code](../images/RL_tamethedog/QRCode.png){:.enlargeImage} -Test everything thoroughly and see if the dog is learning according to its learning rate after the first 10 random responses. Congratulations! You have coded your first AI Reinforcement Learning (RL) project. +Test everything thoroughly and see if the dog is learning according to its learning rate after the first 10 random responses. Be sure to try different learning rates (for example: 0.00, 0.25, 0.50, 0.75, 1.0)for the dog and see that the learning results are in accordance with these different learning rate values. + +Congratulations! You have coded your first AI Reinforcement Learning (RL) project. # Expand Your App