If you have the Watson Studio app, when you create a project, you can choose the project starter that fits your needs. The project starter you choose affects the type of analytic assets you see on the Assets page in the project and the Lite editions of IBM Cloud services that are automatically provisioned. All project starters include data assets and the IBM Cloud Object Storage service to store project files. After you create a project, you can add other analytical assets and associate other services.
When you create a project, you can choose from these project starters:
- Standard: Add collaborators and data assets. Add services as you need them.
- Import project: Import a project from a file.
- Business Analytics: Visualize and share data in dashboards without coding.
- Data Engineering: Cleanse and shape data with a Data Refinery flow.
- Data Science: Analyze data with Jupyter notebooks or RStudio.
- Deep Learning: Develop neural networks and test them in deep learning experiments.
- Modeler: Build, train, test, and deploy machine learning models.
- Streams Flow: Ingest streaming data with a streams flow.
- Visual Recognition: Analyze and classify images in models.
When you create a project in Watson Knowledge Catalog, your projects have the same configuration as the Data Engineering starter.
The Standard project starter provides these capabilities:
- Add and manage collaborators.
- Store files for the project in an IBM Cloud Object Storage bucket.
- Add data assets from connections or by uploading files.
- Integrate with catalogs to work with assets from a catalog or publish assets to a catalog.
- Add any type of analytical asset.
- Associate the required services with the project as you need them.
- The Data Refinery tool to cleanse and shape data.
The import project starter enables importing project assets to use in a new project. You can only import a project that was exported from Watson Studio. You can import a project:
- From a file
You can only import a project that was previously exported from Watson Studio. You cannot manually add other assets to the exported project before importing it.
The Business Analytics project starter provides basic capabilities plus what you need to create and share dashboards:
- The Cognos Dashboard Embedded service to create and publish dashboards.
- A dashboard editor to create dashboards.
See Analytic Dashboard.
The Data Engineering project starter provides basic capabilities plus what you need to refine data:
- The Data Refinery tool to cleanse and shape data.
See Data Refinery.
The Deep Learning project starter provides basic capabilities plus what you need for deep learning experiments:
- A Watson Machine Learning service to create and train models and IBM Cloud integration to deploy models.
- Neural Network modeler to visually design neural networks.
- Experiment Builder tool to define training runs for your experiment and automatically optimize hyperparameters.
- The IBM Cloud command line interface to initiate experiments and monitor training runs.
- Python 3.5 and the Watson Machine Learning library to initiate experiments and monitor your training runs using your preferred coding environment.
See Experiment Builder.
The Data Science project starter provides basic capabilities plus everything you need to run notebooks or RStudio:
- A Jupyter notebook editor for Python, R, or Scala notebooks.
- Rstudio within Watson Studio, as an alternative to running R notebooks.
- A runtime environment that fits your needs:
- A default Anaconda environment
- A Spark environment
- An IBM Apache Spark service
- An Amazon EMR service, if you have Watson Studio Enterprise
- An IBM Analytics Engine service, if you need a Spark cluster
- Visualization libraries and tools to help you tell a story with your data. Watson Studio includes open source libraries like Brunel and PixieDust, as well as the Modeler visualization tool.
- The Decision Optimization engine for running prescriptive analytic APIs within notebooks.
- Open source libraries and packages that provide computation, analytics, and visualization methods. Watson Studio includes some popular open source libraries, such as PySpark, matplotlib, and SparkML.
- A Python API to manipulate data within a notebook.
- Git integration to publish notebooks.
The Modeler project starter provides basic capabilities plus what you need for machine learning models:
- A Watson Machine Learning service to train models and IBM Cloud integration to deploy models.
- A model builder that guides you through building the model step by step.
- A Flow Editor to create a graphical representation of a model.
- SPSS machine learning algorithms and open source machine learning APIs for predictive analytics.
The Streams Flow project starter provides basic capabilities plus what you need for ingesting streaming data:
- The IBM Streaming Analytics service to ingest, analyze, monitor, and correlate data from real-time data sources.
- The streams flow canvas to create a graphical representation of a streams flow.
The Visual Recognition project starter provides basic capabilities plus what you need for visual recognition models:
- A IBM Watson Visual Recognition service to classify images.
- The Visual Recognition model builder to create and train models.
See Visual Recognition.