We are very happy to announce that SPSS Modeler in Watson Studio is now Generally Available! This is great news whether you are a long time SPSS user or a data scientist looking to speed up productivity; or somewhere in between.
Watson Studio Modeler flows provide an interactive environment where you can quickly build machine learning pipelines that flow data from ingestion to transformations and model building; without needing any code. Here is an example flow that builds a decision tree and clusters a popular data set found in the Watson Studio community:
This tool is perfect for rapidly iterating on data sets for finding the right algorithm or testing the right feature engineering. Once you have a model-ready data set, drop one of the 20+ algorithms on your flow to quickly train a model. The branching and experimentation of flows makes it easy to create reproducible research that can be easily understood by any member of your team.
Major benefits of Modeler in Watson Studio
- Built on SPSS Modeler Server — existing Modeler *.str files are supported for import/export, nodes and configurations remain in tact to support existing SPSS Modeler users
- Modeler flows live in projects — Watson Studio projects just got better with Modeler flows now available for use alongside notebooks and other tools
- Updated UI — we kept the backend for all its benefits and updated the UI to be more intuitive and shine in Watson Studio
- One-click model deployment — leverage Watson Machine Learning inside your Modeler flows to save your models for deployment with a single
- Explore data with Data Refinery — previewing data in Modeler flows just got better with the integration of Data Refinery (shown in figure below)
- Advanced model visualization — Modeler flows support viewing models for all available algorithms to help you quickly evaluate model performance and understand how the model fits the data (shown in the figure below)