Machine learning & AI in Watson Studio

IBM Watson Studio is a collaborative environment with AI tools that you and your team can use to collect and prepare training data, and to design, train, and deploy machine learning models.

Ranging from graphical tools you can use to build a model in minutes, to tools that automate running thousands of experiment training runs and hyperparameter optimization, Watson Studio AI tools support popular frameworks, including: TensorFlow, Caffe, PyTorch, and Keras.

You can think of Watson Studio AI tools in three categories:

  • Visual recognition
  • Natural language classification
  • Watson Machine Learning

 

Visual recognition

Tools focused specifically on visual recognition:

  • Built-in models you can use to list keywords describing a given image, or analyze faces in images for age range and gender.
  • A model builder makes it quick and easy to train a model to classify images according to classes you define.
  • Core ML support for using your visual recognition custom models in iOS apps.

See: Visual recognition

 

Natural language classification

Tools focused specifically on natural language classification:

  • A model builder makes it quick and easy to train a model to classify text according to classes you define.
  • API for classifying text in notebooks or in apps you develop.

See: Natural language classification

 

Watson Machine Learning

Tools for designing, training, and managing models:

  • Model builder guides you, step by step, through building a model that uses Spark ML algorithms.
  • AutoAI automatically analyzes your data and generates candidate model pipelines customized for your predictive modeling problem. .
  • SparkML Modeler uses the graphical flow editor so that you can combine nodes and actions to create a machine learning flow.
  • SPSS Modeler uses the SPSS flow editor so that you can quickly develop predictive models using business expertise and deploy them into business operations to improve decision making.
  • Neural Network Modeler presents a graphical view of your model while you build it by combining nodes representing neural network nodes and actions.
  • Experiment builder automates running hundreds of training runs while tracking and storing results.
  • Notebooks provide an interactive programming environment for working with data, testing models, and rapid prototyping.
  • Machine learning command line interface lets you build and work with models in your local environment.

See: Watson Machine Learning