Code details | Deep Learning with R

Code 2

Wed, 29 Jan 2025 00:00:00 +0000

MNIST classification

Sequential neural network for MNIST handwritten digits classification, batch size, learning rate, callbacks, regularization, model capacity, hyperparameter tuning.

Download 02-starter-recipe_code.Rmd. Modified from dl-keras-tf.
Download 02-starter-recipe_images.zip. Extract the images folder in the same directory as the 02-starter-recipe_code.Rmd file.
Read and run the code in the 02-starter-recipe_code.Rmd file. Consult ChatGPT in case of errors.
Answer questions in the code chunks labeled “Your turn!” and describe your observations.
Knit the document into an HTML format and submit the knitted document.

Code 3

Wed, 29 Jan 2025 00:00:00 +0000

Deep learning regression exercise

Download 03_Ames_house_price.Rmd. Modified from dl-keras-tf.
Read and run the code in the 03_Ames_house_price.Rmd file. Consult ChatGPT in case of errors.
Set eval = TRUE and answer questions in the code chunks labeled “Question:” and describe your observations.
Knit the document into an HTML format and submit the knitted document.

Code 1

Mon, 27 Jan 2025 00:00:00 +0000

Self-study

Neural network introduction, main ingredients, activation function, gradient descent, learning rate, ReLU, train-test-validation, prediction

Follow installation instructions in installation_code.R. Note we are installing the keras3 package.
Download 01-main-ingredients_code.Rmd. Modified from dl-keras-tf.
Download 01-main-ingredients_images.zip. Extract the images folder in the same directory as the 01-main-ingredients_code.Rmd file.
Read and run the code in the 01-main-ingredients_code.Rmd file. Consult ChatGPT in case of errors.
Answer questions in the code chunks currently set to eval=FALSE.
Knit the document into an HTML format and submit the knitted document.

Code 5

Mon, 01 Jan 0001 00:00:00 +0000

Computer Vision: MNIST revisted as a CNN

Computer Vision: MNIST revisted as a CNN

Convolutional neural network for MNIST handwritten digits classification.

Download ch08_1_MNIST_classification_CNN_code.Rmd. Modified from dl-keras-tf.
Download ch08_1_MNIST_classification_CNN_images.zip. Extract the images folder in the same directory as the ch08_1_MNIST_classification_CNN_code.Rmd file.
Read and run the code in the ch08_1_MNIST_classification_CNN_code.Rmd file. Consult ChatGPT in case of errors.
Answer questions in the code chunks labeled “Questions” and describe your observations.
Knit the document into an HTML format and submit the knitted document.