{"id":8035,"date":"2023-10-25T15:19:34","date_gmt":"2023-10-25T23:19:34","guid":{"rendered":"https:\/\/live-cometml.pantheonsite.io\/?p=8035"},"modified":"2025-04-24T17:05:08","modified_gmt":"2025-04-24T17:05:08","slug":"text-classification-using-r-keras-and-comet-ml","status":"publish","type":"post","link":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml\/","title":{"rendered":"Text Classification Using R, Keras, and Comet ML"},"content":{"rendered":"\n\n\n\n\n
\"\"\/
Source: https:\/\/unsplash.com\/photos\/IhWYiwSxm8g<\/a><\/figcaption><\/figure>\n\n\n\n

Text classification is an interesting application of natural language processing. It is a supervised learning methodology that predicts if a piece of text belongs to one category or the other. As a machine learning engineer, you start with a labeled data set that has vast amounts of text that have already been categorized. These algorithms can perform sentiment analysis, create spam filters, and other applications.<\/p>\n\n\n\n

This tutorial will teach you how to train your binary text classifiers using Keras. You will use the IMDB datase<\/a>t that has 50K+ movie reviews that are classified as positive or negative. The main goal of this tutorial is to equip you with the skills of using R and R Studio to build the classifier while using Comet ML\u2019s platform to monitor your experiments.<\/p>\n\n\n\n

Pre-requisites<\/strong><\/h2>\n\n\n\n

To follow along with this tutorial, you need to prepare your development environment as follows:<\/p>\n\n\n\n

\u00b7 Make sure you have R installed from the Official Website<\/a><\/p>\n\n\n\n

\u00b7 After installing R, install R Studio<\/a>, the preferred IDE for creating R projects.<\/p>\n\n\n\n

\u00b7 Visit Comet ML\u2019s platform<\/a> to sign up for an account<\/p>\n\n\n\n

Getting Started<\/strong><\/h2>\n\n\n\n

Let\u2019s install some dependencies that you need to build your classifier. You will install all these dependencies by using R Studio. Open R Studio and in the console, type in the following commands to install the dependencies needed.<\/p>\n\n\n\n

First, install Comet\u2019s R package which will be used to log metrics to your Comet<\/p>\n\n\n\n

install.packages(<\/span>\u201ccometr\u201d)<\/span><\/span><\/pre>\n\n\n\n
Next, you will need to install the Keras package for R by running the following command in the console. Keras provides a simple API to build neural networks and uses Tensorflow as the backend.<\/p>\n\n\n\n
install.packages(<\/span>\u201ckeras\u201d)<\/span>\ninstall_keras(<\/span>)<\/span><\/span><\/pre>\n\n\n\n
Now we have our development environment setup. Let\u2019s move on to loading the datasets.<\/p>\n\n\n\n
Data Loading<\/strong><\/h2>\n\n\n\n
The IMDB dataset is already preloaded with Keras, making it easy to load it. Create a new R Script and call it train.R. This file will hold all your source code for this project. Add the code below to download the IMDB dataset that has 50K+ reviews for movies from the IMDB website.<\/p>\n\n\n\n
library(<\/span>keras)<\/span>\nlibrary(<\/span>cometr)<\/span>\nimdb <-<\/span> dataset_imdb(<\/span>num_words =<\/span> 10000<\/span>)<\/span>\ntrain_data <-<\/span> imdb$<\/span>train$<\/span>x\ntrain_labels <-<\/span> imdb$<\/span>train$<\/span>y\ntest_data <-<\/span> imdb$<\/span>test$<\/span>x\ntest_labels <-<\/span> imdb$<\/span>test$<\/span>y<\/span><\/pre>\n\n\n\n
The data is partially preprocessed. This means that the sequence of words representing reviews has already been converted to a sequence of integers. Each integer stands for a unique word in a dictionary.<\/p>\n\n\n\n
However, you will still need to convert the sequence of numbers to vectors. In order to do this, you will write a simple function that vectorizes your lists. This function will use the one-hot encoding methodology. The one-hot encoding converts your lists of integers to an N-dimensional vector where N is the count of unique words in the data. The term being represented will have a value of 1 while the rest are represented by 0. You can choose to explore other vectorization functions like TF-IDF that are a bit more memory efficient.<\/p>\n\n\n\n
vectorize_seq <-<\/span> function<\/span>(<\/span>sequences,<\/span> dimension =<\/span> 10000<\/span>)<\/span> {<\/span>\n  results <-<\/span> matrix(<\/span>0<\/span>,<\/span> nrow =<\/span> length<\/span>(<\/span>sequences)<\/span>,<\/span> ncol =<\/span> dimension)<\/span>\n  for<\/span> (<\/span>i in<\/span> 1<\/span>:<\/span>length<\/span>(<\/span>sequences)<\/span>)<\/span>\n    results[<\/span>i,<\/span> sequences[[<\/span>i]<\/span>]<\/span>]<\/span> <-<\/span> 1<\/span>\n  results\n}<\/span><\/span><\/pre>\n\n\n\n
Make sure you convert the training and testing datasets using the function created above.<\/p>\n\n\n\n
x_train <-<\/span> vectorize_seq(<\/span>train_data)<\/span>\nx_test <-<\/span> vectorize_seq(<\/span>test_data)<\/span>\n\ny_train <-<\/span> as.numeric<\/span>(<\/span>train_labels)<\/span>\ny_test <-<\/span> as.numeric<\/span>(<\/span>test_labels)<\/span><\/span><\/pre>\n\n\n\n
Training Neural Network<\/code><\/h2>\n\n\n\n
Let\u2019s use a simple architecture to build the network. A stack of fully connected layers with relu activation functions can solve our binary classifications problem. However, some design decisions need to be made<\/p>\n\n\n\n
\u00b7 How many layers will your network have?<\/p>\n\n\n\n
\u00b7 How many hidden units do you need for each layer?<\/p>\n\n\n\n
For this tutorial, let\u2019s keep it simple and have the architecture as shown below.<\/p>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
It is also important to note that you will need a sigmoid function that will be used to calculate the output probabilities (0 for negative or 1 for positive). Keras provides a simple API and you can create your model as shown in the code block below.<\/p>\n\n\n\n
model <-<\/span> keras_model_sequential(<\/span>)<\/span> %>%<\/span>\n  layer_dense(<\/span>units =<\/span> 16<\/span>,<\/span> activation =<\/span> \"relu\"<\/span>,<\/span> input_shape =<\/span> c<\/span>(<\/span>10000<\/span>)<\/span>)<\/span> %>%<\/span>\n  layer_dense(<\/span>units =<\/span> 16<\/span>,<\/span> activation =<\/span> \"relu\"<\/span>)<\/span> %>%<\/span>\n  layer_dense(<\/span>units =<\/span> 1<\/span>,<\/span> activation =<\/span> \"sigmoid\"<\/span>)<\/span><\/span><\/pre>\n\n\n\n
Loss Function & Optimizers<\/strong><\/h2>\n\n\n\n
This section demonstrates how to add loss functions and optimizers to your neural network. Because the output of the network is a probability, you can use a binary cross-entropy loss function. You will use a learning rate of 0.01 for your optimizer. Add the following code block to your R script.<\/p>\n\n\n\n
learning_rate <-<\/span> 0.001\nmodel %>%<\/span> compile(<\/span>\n  optimizer =<\/span> optimizer_rmsprop(<\/span>lr=<\/span>learning_rate)<\/span>,<\/span>\n  loss =<\/span> \"binary_crossentropy\"<\/span>,<\/span>\n  metrics =<\/span> c<\/span>(<\/span>\"accuracy\"<\/span>)<\/span>\n)<\/span> <\/span><\/pre>\n\n\n\n
Monitoring<\/strong><\/h2>\n\n\n\n
Before you start training, you\u2019ll need to integrate Comet ML\u2019s R package by adding the following code block. Make sure you have your API key from your Comet ML\u2019s account then create a ~\/.comet.yml.<\/code> Follow the official documentation<\/a> for additional help with getting started with R.<\/p>\n\n\n\n
COMET_WORKSPACE:<\/span> my-comet-user-name<\/span>\nCOMET_PROJECT_NAME:<\/span> nlpkeras<\/span>\nCOMET_API_KEY:<\/span> my-comet-api-key<\/span><\/span><\/pre>\n\n\n\n
Next, create an experiment as shown below in your R script.<\/p>\n\n\n\n
exp<\/span> <-<\/span> create_experiment(<\/span>\n  keep_active =<\/span> TRUE<\/span>,<\/span>\n  log_output =<\/span> TRUE<\/span>,<\/span>\n  log_error =<\/span> FALSE<\/span>,<\/span>\n  log_code =<\/span> TRUE<\/span>,<\/span>\n  log_system_details =<\/span> TRUE<\/span>,<\/span>\n  log_git_info =<\/span> TRUE<\/span>\n)<\/span><\/span><\/pre>\n\n\n\n
Next, you need to specify the number of epochs you want to train the network and log it using Comet ML.<\/p>\n\n\n\n
epochs <-<\/span> 20\nexp<\/span>$<\/span>log_parameter(<\/span>\"epochs\"<\/span>,<\/span> epochs)<\/span>\nexp<\/span>$<\/span>log_parameter(<\/span>\"learning_rate\"<\/span>,<\/span> learning_rate)<\/span><\/span><\/pre>\n\n\n\n
Training<\/strong><\/h2>\n\n\n\n
Now, train the model using the training datasets created above. Also, make sure you add the model evaluation using the validation sets created above.<\/p>\n\n\n\n
history <-<\/span> model %>%<\/span> fit(<\/span>\n  partial_x_train,<\/span>\n  partial_y_train,<\/span>\n  epochs =<\/span> epochs,<\/span>\n  batch_size =<\/span> 512<\/span>,<\/span>\n  validation_data =<\/span> list<\/span>(<\/span>x_val,<\/span> y_val)<\/span>\n)<\/span><\/span><\/pre>\n\n\n\n
The fit method returns a history object that has a plot function that you can use to visualize the training and validation metrics by epoch.<\/p>\n\n\n\n
plot(<\/span>history)<\/span><\/span><\/pre>\n\n\n\n
Make sure you log the training loss and accuracy metrics to Comet ML.<\/p>\n\n\n\n
results <-<\/span> model %>%<\/span> evaluate(<\/span>x_test,<\/span> y_test)<\/span>\nexp<\/span>$<\/span>log_metric(<\/span>\"val_loss\"<\/span>,<\/span> results$<\/span>loss)<\/span>\nexp<\/span>$<\/span>log_metric(<\/span>\"val_acc\"<\/span>,<\/span> results$<\/span>acc)<\/span><\/span><\/pre>\n\n\n\n
Here\u2019s an example of visualization from Comet ML\u2019s platform showcasing the model\u2019s accuracy.<\/p>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
Conclusion<\/h2>\n\n\n\n
In this article we used Keras with R to build a neural network that can classify text into two categories. In addition, we logged some metrics like loss, accuracy, and epochs to Comet ML\u2019s platform. This tutorial was just a simple introduction to how to use R to build NLP models while monitoring your experiments using Comet ML. Kindly visit Comet ML\u2019s Official Documentation<\/a>to gain more insights on how to monitor your R projects.<\/p>\n","protected":false},"excerpt":{"rendered":"
Text classification is an interesting application of natural language processing. It is a supervised learning methodology that predicts if a piece of text belongs to one category or the other. As a machine learning engineer, you start with a labeled data set that has vast amounts of text that have already been categorized. These algorithms […]<\/p>\n","protected":false},"author":65,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"customer_name":"","customer_description":"","customer_industry":"","customer_technologies":"","customer_logo":"","footnotes":""},"categories":[9,7],"tags":[],"coauthors":[165],"class_list":["post-8035","post","type-post","status-publish","format-standard","hentry","category-product","category-tutorials"],"yoast_head":"\nText Classification Using R, Keras, and Comet ML - Comet<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Text Classification Using R, Keras, and Comet ML\" \/>\n<meta property=\"og:description\" content=\"Text classification is an interesting application of natural language processing. It is a supervised learning methodology that predicts if a piece of text belongs to one category or the other. As a machine learning engineer, you start with a labeled data set that has vast amounts of text that have already been categorized. These algorithms […]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml\" \/>\n<meta property=\"og:site_name\" content=\"Comet\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/cometdotml\" \/>\n<meta property=\"article:published_time\" content=\"2023-10-25T23:19:34+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2025-04-24T17:05:08+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*V9y2IkTtXlovTPiHESe-Cw.jpeg\" \/>\n<meta name=\"author\" content=\"Klurdy Studios\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@Cometml\" \/>\n<meta name=\"twitter:site\" content=\"@Cometml\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Klurdy Studios\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"6 minutes\" \/>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Text Classification Using R, Keras, and Comet ML - Comet","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml","og_locale":"en_US","og_type":"article","og_title":"Text Classification Using R, Keras, and Comet ML","og_description":"Text classification is an interesting application of natural language processing. It is a supervised learning methodology that predicts if a piece of text belongs to one category or the other. As a machine learning engineer, you start with a labeled data set that has vast amounts of text that have already been categorized. These algorithms […]","og_url":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml","og_site_name":"Comet","article_publisher":"https:\/\/www.facebook.com\/cometdotml","article_published_time":"2023-10-25T23:19:34+00:00","article_modified_time":"2025-04-24T17:05:08+00:00","og_image":[{"url":"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*V9y2IkTtXlovTPiHESe-Cw.jpeg","type":"","width":"","height":""}],"author":"Klurdy Studios","twitter_card":"summary_large_image","twitter_creator":"@Cometml","twitter_site":"@Cometml","twitter_misc":{"Written by":"Klurdy Studios","Est. reading time":"6 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml#article","isPartOf":{"@id":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml\/"},"author":{"name":"Klurdy Studios","@id":"https:\/\/www.comet.com\/site\/#\/schema\/person\/069e186ad4a5b6d6950292821ea0f37b"},"headline":"Text Classification Using R, Keras, and Comet ML","datePublished":"2023-10-25T23:19:34+00:00","dateModified":"2025-04-24T17:05:08+00:00","mainEntityOfPage":{"@id":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml\/"},"wordCount":897,"publisher":{"@id":"https:\/\/www.comet.com\/site\/#organization"},"image":{"@id":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml#primaryimage"},"thumbnailUrl":"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*V9y2IkTtXlovTPiHESe-Cw.jpeg","articleSection":["Product","Tutorials"],"inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml\/","url":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml","name":"Text Classification Using R, Keras, and Comet ML - Comet","isPartOf":{"@id":"https:\/\/www.comet.com\/site\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml#primaryimage"},"image":{"@id":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml#primaryimage"},"thumbnailUrl":"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*V9y2IkTtXlovTPiHESe-Cw.jpeg","datePublished":"2023-10-25T23:19:34+00:00","dateModified":"2025-04-24T17:05:08+00:00","breadcrumb":{"@id":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml#primaryimage","url":"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*V9y2IkTtXlovTPiHESe-Cw.jpeg","contentUrl":"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*V9y2IkTtXlovTPiHESe-Cw.jpeg"},{"@type":"BreadcrumbList","@id":"https:\/\/www.comet.com\/site\/blog\/text-classification-using-r-keras-and-comet-ml#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.comet.com\/site\/"},{"@type":"ListItem","position":2,"name":"Text Classification Using R, Keras, and Comet ML"}]},{"@type":"WebSite","@id":"https:\/\/www.comet.com\/site\/#website","url":"https:\/\/www.comet.com\/site\/","name":"Comet","description":"Build Better Models Faster","publisher":{"@id":"https:\/\/www.comet.com\/site\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.comet.com\/site\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/www.comet.com\/site\/#organization","name":"Comet ML, Inc.","alternateName":"Comet","url":"https:\/\/www.comet.com\/site\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.comet.com\/site\/#\/schema\/logo\/image\/","url":"https:\/\/www.comet.com\/site\/wp-content\/uploads\/2025\/01\/logo_comet_square.png","contentUrl":"https:\/\/www.comet.com\/site\/wp-content\/uploads\/2025\/01\/logo_comet_square.png","width":310,"height":310,"caption":"Comet ML, Inc."},"image":{"@id":"https:\/\/www.comet.com\/site\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/cometdotml","https:\/\/x.com\/Cometml","https:\/\/www.youtube.com\/channel\/UCmN63HKvfXSCS-UwVwmK8Hw"]},{"@type":"Person","@id":"https:\/\/www.comet.com\/site\/#\/schema\/person\/069e186ad4a5b6d6950292821ea0f37b","name":"Klurdy Studios","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.comet.com\/site\/#\/schema\/person\/image\/b1a3bf3caaa793aaad2da005b3ba38ba","url":"https:\/\/www.comet.com\/site\/wp-content\/uploads\/2023\/08\/1635710213869-96x96.jpg","contentUrl":"https:\/\/www.comet.com\/site\/wp-content\/uploads\/2023\/08\/1635710213869-96x96.jpg","caption":"Klurdy Studios"},"url":"https:\/\/www.comet.com\/site\/blog\/author\/brianklurdy-com\/"}]}},"_links":{"self":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts\/8035","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/users\/65"}],"replies":[{"embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/comments?post=8035"}],"version-history":[{"count":1,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts\/8035\/revisions"}],"predecessor-version":[{"id":15483,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts\/8035\/revisions\/15483"}],"wp:attachment":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/media?parent=8035"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/categories?post=8035"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/tags?post=8035"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/coauthors?post=8035"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}