h2o_deeplearning_gridsearch_mnist_example.R

library(h2o)
h2o.init(nthreads = -1)  # This means nthreads = num available cores

train_file <- "https://h2o-public-test-data.s3.amazonaws.com/bigdata/laptop/mnist/train.csv.gz"
test_file <- "https://h2o-public-test-data.s3.amazonaws.com/bigdata/laptop/mnist/test.csv.gz"

train <- h2o.importFile(train_file)
test <- h2o.importFile(test_file)

# To see a brief summary of the data, run the following command
summary(train)
summary(test)


# Specify the response and predictor columns
y <- "C785"
x <- setdiff(names(train), y)

# We encode the response column as categorical for multinomial classification
train[,y] <- as.factor(train[,y])
test[,y] <- as.factor(test[,y])

# Train a Deep Learning model and validate on a test set
model <- h2o.deeplearning(
        x = x, 
        y = y, 
        training_frame = train,
        validation_frame = test,   
        distribution = "multinomial",
        activation = "RectifierWithDropout", 
        hidden = c(200,200,200), 
        input_dropout_ratio = 0.2, 
        l1 = 1e-5, 
        epochs = 10)


hidden_opt <- list(c(200,200), c(100,300,100), c(500,500,500))
l1_opt <- c(1e-5,1e-7)
hyper_params <- list(hidden = hidden_opt, l1 = l1_opt)

model_grid <- h2o.grid(
        "deeplearning", 
        hyper_params = hyper_params, 
        x = x,
        y = y,
        distribution = "multinomial", 
        training_frame = train, 
        validation_frame = test)


# print out all prediction errors and run times of the models
model_grid

# print out the Test MSE for all of the models
for (model_id in model_grid@model_ids) {
  model <- h2o.getModel(model_id)
  mse <- h2o.mse(model, valid = TRUE)
  print(sprintf("Test set MSE: %f", mse))
}