Autoencoder in R

packages

# Helper packages
library(dplyr)    # for data manipulation
library(ggplot2)  # for data visualization

# Modeling packages
library(h2o)  # for fitting autoencoders

train data

mnist <- dslabs::read_mnist()
names(mnist)

initialize H2O

h2o.no_progress()  # turn off progress bars
h2o.init(max_mem_size = "5g")  # initialize H2O instance

H2O is not running yet, starting it now...

Note:  In case of errors look at the following log files:
    C:\Users\zouhu\AppData\Local\Temp\Rtmp4APVUM\file521043104b5e/h2o_zouhua_started_from_r.out
    C:\Users\zouhu\AppData\Local\Temp\Rtmp4APVUM\file521021935c59/h2o_zouhua_started_from_r.err

java version "1.8.0_231"
Java(TM) SE Runtime Environment (build 1.8.0_231-b11)
Java HotSpot(TM) 64-Bit Server VM (build 25.231-b11, mixed mode)

Starting H2O JVM and connecting: . Connection successful!

R is connected to the H2O cluster: 
    H2O cluster uptime:         5 seconds 583 milliseconds 
    H2O cluster timezone:       Asia/Shanghai 
    H2O data parsing timezone:  UTC 
    H2O cluster version:        3.32.0.1 
    H2O cluster version age:    2 months and 30 days  
    H2O cluster name:           H2O_started_from_R_zouhua_vrf494 
    H2O cluster total nodes:    1 
    H2O cluster total memory:   4.44 GB 
    H2O cluster total cores:    6 
    H2O cluster allowed cores:  6 
    H2O cluster healthy:        TRUE 
    H2O Connection ip:          localhost 
    H2O Connection port:        54321 
    H2O Connection proxy:       NA 
    H2O Internal Security:      FALSE 
    H2O API Extensions:         Amazon S3, Algos, AutoML, Core V3, TargetEncoder, Core V4 
    R Version:                  R version 4.0.2 (2020-06-22) 

Comparing PCA to an autoencoder

# Convert mnist features to an h2o input data set
features <- as.h2o(mnist$train$images)

# Train an autoencoder
ae1 <- h2o.deeplearning(
  x = seq_along(features),
  training_frame = features,
  autoencoder = TRUE,
  hidden = 2,
  activation = 'Tanh',
  sparse = TRUE
)

# Extract the deep features
ae1_codings <- h2o.deepfeatures(ae1, features, layer = 1)
ae1_codings

Stacked autoencoders

# Hyperparameter search grid
hyper_grid <- list(hidden = list(
  c(50),
  c(100), 
  c(300, 100, 300),
  c(100, 50, 100),
  c(250, 100, 50, 100, 250)
))

# Execute grid search
ae_grid <- h2o.grid(
  algorithm = 'deeplearning',
  x = seq_along(features),
  training_frame = features,
  grid_id = 'autoencoder_grid',
  autoencoder = TRUE,
  activation = 'Tanh',
  hyper_params = hyper_grid,
  sparse = TRUE,
  ignore_const_cols = FALSE,
  seed = 123
)

# Print grid details
h2o.getGrid('autoencoder_grid', sort_by = 'mse', decreasing = FALSE)

visualization

# Get sampled test images
index <- sample(1:nrow(mnist$test$images), 4)
sampled_digits <- mnist$test$images[index, ]
colnames(sampled_digits) <- paste0("V", seq_len(ncol(sampled_digits)))

# Predict reconstructed pixel values
best_model_id <- ae_grid@model_ids[[1]]
best_model <- h2o.getModel(best_model_id)
reconstructed_digits <- predict(best_model, as.h2o(sampled_digits))
names(reconstructed_digits) <- paste0("V", seq_len(ncol(reconstructed_digits)))

combine <- rbind(sampled_digits, as.matrix(reconstructed_digits))

# Plot original versus reconstructed
par(mfrow = c(1, 3), mar=c(1, 1, 1, 1))
layout(matrix(seq_len(nrow(combine)), 4, 2, byrow = FALSE))
for(i in seq_len(nrow(combine))) {
  image(matrix(combine[i, ], 28, 28)[, 28:1], xaxt="n", yaxt="n")
}

Reference

1. Autoencoder